{"id":72712,"date":"2022-09-30T14:02:01","date_gmt":"2022-09-30T14:02:01","guid":{"rendered":"https:\/\/loopbeatzz.info\/wp\/fuzzy-clustering-and-fuzzy-network-based-interpretable-fuzzy-model-for-prediction-scientific-reports-nature-com\/"},"modified":"2022-09-30T14:02:01","modified_gmt":"2022-09-30T14:02:01","slug":"fuzzy-clustering-and-fuzzy-network-based-interpretable-fuzzy-model-for-prediction-scientific-reports-nature-com","status":"publish","type":"post","link":"https:\/\/loopbeatzz.info\/wp\/fuzzy-clustering-and-fuzzy-network-based-interpretable-fuzzy-model-for-prediction-scientific-reports-nature-com\/","title":{"rendered":"Fuzzy-clustering and fuzzy network based interpretable fuzzy model for prediction | Scientific Reports &#8211; Nature.com"},"content":{"rendered":"<p>Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain             the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in             Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles             and JavaScript.<br \/>Advertisement<br \/>                              <a data-test=\"journal-link\" href=\"\/srep\"><i data-test=\"journal-title\">Scientific Reports<\/i><\/a>                          <b data-test=\"journal-volume\"><span class=\"u-visually-hidden\">volume<\/span>\u00a012<\/b>, Article\u00a0number:\u00a0<span data-test=\"article-number\">16279<\/span> (<span data-test=\"article-publication-year\">2022<\/span>)             <a href=\"#citeas\" class=\"c-article-info-details__cite-as u-hide-print\" data-track=\"click\" data-track-action=\"cite this article\" data-track-label=\"link\">Cite this article<\/a>                     <br \/><a href=\"\/articles\/s41598-022-20015-y\/metrics\" data-track=\"click\" data-track-action=\"view metrics\" data-track-label=\"link\" rel=\"nofollow\">Metrics <span class=\"u-visually-hidden\">details<\/span><\/a><br \/>Interpretability is the dominant feature of a fuzzy model in security-oriented fields. Traditionally fuzzy models based on expert knowledge have obtained well interpretation innately but imprecisely. Numerical data based fuzzy models perform well in precision but not necessarily in interpretation. To utilize the expert knowledge and numerical data in a fuzzy model synchronously, this paper proposed a hybrid fuzzy c-means (FCM) clustering algorithm and Fuzzy Network (FN) method-based model for prediction. The Mamdani rule-based structure of the proposed model is identified based on FCM algorithm from data and by expert-system method from expert knowledge, both of which are combined by FN method. Particle swarm optimization (PSO) algorithm is utilized to optimize the fuzzy set parameters. We tested the proposed model on 6 real datasets comparing the results with the ones obtained by using FCM algorithm. The results showed that our model performed best in interpretability, transparency, and accuracy.<br \/>Fuzzy model is to implement logical reasoning and intelligent calculation for fuzzy information without a definite mathematical model. The main method is a set of IF\u2013THEN rules-based procession designed either from expert knowledge or numerical data. Although independent on accurate mathematical model, fuzzy model is a powerful technique for logical reasoning, numerical calculation and non-linear function approximation<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 1\" title=\"Wang, G., Wang, H. &amp; Long, Z. Norm approximation of mamdani fuzzy system to a class of integrable functions. Int. J. Fuzzy Syst. 23, 833\u2013848.                    https:\/\/doi.org\/10.1007\/s40815-020-01008-3                                     (2021).\" href=\"\/articles\/s41598-022-20015-y#ref-CR1\" id=\"ref-link-section-d104546385e378\">1<\/a><\/sup>.<br \/>To identify a fuzzy model, structure establishment and parameter estimations are two main steps<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 2\" title=\"Tsekouras, G., Sarimveis, H., Kavakli, E. &amp; Bafas, G. A hierarchical fuzzy-clustering approach to fuzzy modeling. Fuzzy Sets Syst. 150, 245\u2013266.                    https:\/\/doi.org\/10.1016\/j.fss.2004.04.013                                     (2005).\" href=\"\/articles\/s41598-022-20015-y#ref-CR2\" id=\"ref-link-section-d104546385e385\">2<\/a><\/sup>. Structure identification is related to the number of rules after the important input variables have been selected. Parameter estimation (fuzzy set parameters) descripts a reliable non-linear approximation system.<br \/>Interpretability is the major feature of a fuzzy model in special fields, for example in security-oriented fields, and military field. There is a large amount of knowledge on top of the experience of experts. When this kind of knowledge is dealt with by machine learning, the security of this fuzzy logic (FL) model is vital for military application, and also fundamental for the transparency and interpretability of the knowledge. This kind of knowledge expressed by FL is conceptually easy to understand, tolerant of the imprecise information, and easy for human communication. Thus, expert knowledge based fuzzy models are traditionally well interpretable<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 3\" title=\"Zhang, Y., Wu, X. B., Xing, Z. Y. &amp; Hu, W. L. On generating interpretable and precise fuzzy systems based on Pareto multi-objective cooperative co-evolutionary algorithm. Appl. Soft Comput. 11, 1284\u20131294.                    https:\/\/doi.org\/10.1016\/j.asoc.2010.03.005                                     (2011).\" href=\"\/articles\/s41598-022-20015-y#ref-CR3\" id=\"ref-link-section-d104546385e392\">3<\/a><\/sup>, which is the motivation of the approach with fuzzy logic.<br \/>Expert knowledge is mainly expressed in Mamdani-type IF\u2013THEN rules because this structure is more interpretable than T-S-type ones. Arguably, traditionally fuzzy models based on expert knowledge have obtained well interpretable innately. So, Mamdani rule-based structure is suitable for an interpretable-oriented fuzzy model.<br \/>Numerical data based fuzzy models are not necessarily interpretable, but lots of them perform well in precision with the help of soft computing techniques. For example, artificial Neural network (ANN), inspired by neuroscience, is one of the most successful methods in \u201clearning\u201d from data. Modern ANNs, particularly \u201cdeep learning\u201d models, have been sped up by the increase of raw computer power. But \u201cthey cannot approach the cognitive capabilities of a four-year old. Perhaps more striking is that ANNs remain even further from approaching the abilities of simple animals<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 4\" title=\"Zador, A. M. A critique of pure learning and what artificial neural networks can learn from animal brains. Nat. Commun. 10, 1\u20137.                    https:\/\/doi.org\/10.1038\/s41467-019-11786-6                                     (2019).\" href=\"\/articles\/s41598-022-20015-y#ref-CR4\" id=\"ref-link-section-d104546385e403\">4<\/a><\/sup>.\u201d It implies that the innate structure plays the dominant role in the learning capability.<br \/>The FCM clustering algorithm is a powerful unsupervised learning technique to form a few rules with simple and interpretable structure<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 5\" title=\"Dick, S., Meeks, A., Last, M., Bunke, H. &amp; Kandel, A. Data mining in software metrics databases. Fuzzy Sets Syst. 145, 81\u2013110.                    https:\/\/doi.org\/10.1016\/j.fss.2003.10.006                                     (2004).\" href=\"\/articles\/s41598-022-20015-y#ref-CR5\" id=\"ref-link-section-d104546385e410\">5<\/a><\/sup>. FCM induces rules by organizing and categorizing data into partitions. Partitions with homogeneous data form clusters, and each cluster is associated with a rule. The fuzzy sets of rules are independent from each other. Each dimension of data is tailored only for one rule<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 6\" title=\"Guillaume, S. Designing fuzzy inference systems from data: An interpretability-oriented review. Fuzzy Syst. IEEE Trans. 9, 426\u2013443.                    https:\/\/doi.org\/10.1109\/91.928739                                     (2001).\" href=\"\/articles\/s41598-022-20015-y#ref-CR6\" id=\"ref-link-section-d104546385e414\">6<\/a><\/sup>. In order to improve the precision, many optimization-based FCM models have been proposed combining with metaheuristic optimization algorithms, such as genetic algorithm(GA)<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 7\" title=\"Jiang, G. C. &amp; Sun, Z. J. Improvement on a fuzzy C-means algorithm based on genetic algorithm. Appl. Mech. Mater. 614, 385\u2013388.                    https:\/\/doi.org\/10.4028\/www.scientific.net\/AMM.614.385                                     (2014).\" href=\"\/articles\/s41598-022-20015-y#ref-CR7\" id=\"ref-link-section-d104546385e418\">7<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 8\" title=\"Ding, Y. &amp; Fu, X. Kernel-based fuzzy c-means clustering algorithm based on genetic algorithm. Neurocomputing 188, 233\u2013238.                    https:\/\/doi.org\/10.1016\/j.neucom.2015.01.106                                     (2016).\" href=\"\/articles\/s41598-022-20015-y#ref-CR8\" id=\"ref-link-section-d104546385e421\">8<\/a><\/sup> and PSO<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" title=\"Silva-Filho, T. M., Pimentel, B. A., Souza, R. M. C. R. &amp; Oliveira, A. L. I. Hybrid methods for fuzzy clustering based on fuzzy c-means and improved particle swarm optimization. Expert Syst. Appl. 42, 6315\u20136328.                    https:\/\/doi.org\/10.1016\/j.eswa.2015.04.032                                     (2015).\" href=\"#ref-CR9\" id=\"ref-link-section-d104546385e425\">9<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" title=\"Kuo, R. J., Lin, T. C., Zulvia, F. E. &amp; Tsai, C. Y. A hybrid metaheuristic and kernel intuitionistic fuzzy c-means algorithm for cluster analysis. Appl. Soft Comput. 67, 299\u2013308.                    https:\/\/doi.org\/10.1016\/j.asoc.2018.02.039                                     (2018).\" href=\"#ref-CR10\" id=\"ref-link-section-d104546385e425_1\">10<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" title=\"Izakian, H. &amp; Abraham, A. Fuzzy C-means and fuzzy swarm for fuzzy clustering problem. Expert Syst. Appl. 38, 1835\u20131838.                    https:\/\/doi.org\/10.1016\/j.eswa.2010.07.112                                     (2011).\" href=\"#ref-CR11\" id=\"ref-link-section-d104546385e425_2\">11<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" title=\"Hu, J. &amp; Yin, H. FCM clustering algorithm based on PSO-TVAC algorithm with adaptively weighted centers. Adv. Appl. Math. 10, 953\u2013962.                    https:\/\/doi.org\/10.12677\/AAM.2021.104104                                     (2021).\" href=\"#ref-CR12\" id=\"ref-link-section-d104546385e425_3\">12<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" title=\"Zhang, Y. C., Xiong, X. &amp; Zhang, Q. D. An improved self-adaptive PSO algorithm with detection function for multimodal function optimization problems. Math. Probl. Eng. 657\u2013675, 2013.                    https:\/\/doi.org\/10.1155\/2013\/716952                                     (2013).\" href=\"#ref-CR13\" id=\"ref-link-section-d104546385e425_4\">13<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" title=\"Alam, S., Dobbie, G., Koh, Y. S., Riddle, P. &amp; Rehman, S. U. Research on particle swarm optimization based clustering: A systematic review of literature and techniques. Swarm Evol. Comput. 17, 1\u201313.                    https:\/\/doi.org\/10.1016\/j.swevo.2014.02.001                                     (2014).\" href=\"#ref-CR14\" id=\"ref-link-section-d104546385e425_5\">14<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 15\" title=\"Chen, S., Xu, Z. &amp; Tang, Y. A hybrid clustering algorithm based on fuzzy c-means and improved particle swarm optimization. Arab. J. Sci. Eng. 39, 8875\u20138887.                    https:\/\/doi.org\/10.1007\/s13369-014-1424-9                                     (2014).\" href=\"\/articles\/s41598-022-20015-y#ref-CR15\" id=\"ref-link-section-d104546385e428\">15<\/a><\/sup>. Both GA and PSO aim to solve optimization problems without being trapped into local minima. Due to its versatility and simplicity, PSO has become one of the most popular metaheuristics and an important tool for many applications.<br \/>The fuzzy models for prediction should utilize the expert knowledge and numerical data in a fuzzy model synchronously<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 16\" title=\"Ahmed, M. A. &amp; Al-Jamimi, H. A. Machine learning approaches for predicting software maintainability: A fuzzy-based transparent model. IET Softw. 7, 317\u2013326.                    https:\/\/doi.org\/10.1049\/iet-sen.2013.0046                                     (2013).\" href=\"\/articles\/s41598-022-20015-y#ref-CR16\" id=\"ref-link-section-d104546385e435\">16<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 17\" title=\"Wang, L.-X. &amp; Mendel, J. M. Generating fuzzy rules by learning from examples. IEEE Trans. Syst. Man Cybern. 22, 1414\u20131427.                    https:\/\/doi.org\/10.1109\/21.199466                                     (1992).\" href=\"\/articles\/s41598-022-20015-y#ref-CR17\" id=\"ref-link-section-d104546385e438\">17<\/a><\/sup>. Historical data provides numerical quantitative measurements from past projects regarding the internal and external quality attributes. Experts utilize their experience to provide \u201cfuzzy\u201d information, or qualitative descriptions of the correlation between the internal and external quality attributes<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 18\" title=\"Wang, X., Gegov, A., Farzad, A., Chen, Y. &amp; Hu, Q. Fuzzy network based framework for software maintainability prediction. Internat. J. Uncertain. Fuzziness Knowl.-Based Syst. 27, 841\u2013862.                    https:\/\/doi.org\/10.1142\/s0218488519500375                                     (2019).\" href=\"\/articles\/s41598-022-20015-y#ref-CR18\" id=\"ref-link-section-d104546385e442\">18<\/a><\/sup>. To make the model for prediction more practical, interpretability is crucial as it allows the practitioners to provide their own judgment on the predictors in linguistic terms. Black box prediction models (e.g., ANN-based models), are hard to identify the structure and to incorporate the experts\u2019 judgments.<br \/>The novel method of FN is capable to combine the rules both from knowledge and from data. As a kind of Chained Fuzzy System (CFS)<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 19\" title=\"Bucolo, M., Fortuna, L. &amp; Rosa, M. L. Complex dynamics through fuzzy chains. IEEE Trans. Fuzzy Syst. 12, 289\u2013295.                    https:\/\/doi.org\/10.1109\/TFUZZ.2004.825969                                     (2004).\" href=\"\/articles\/s41598-022-20015-y#ref-CR19\" id=\"ref-link-section-d104546385e449\">19<\/a><\/sup> or Hierarchical Fuzzy System (HFS)<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 20\" title=\"Lee, M.-L., Chung, H.-Y. &amp; Yu, F.-M. Modeling of hierarchical fuzzy systems. Fuzzy Sets Syst. 138, 343\u2013361.                    https:\/\/doi.org\/10.1016\/S0165-0114(02)00517-1                                     (2003).\" href=\"\/articles\/s41598-022-20015-y#ref-CR20\" id=\"ref-link-section-d104546385e453\">20<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 21\" title=\"Joo, M. G. &amp; Lee, J. S. A class of hierarchical fuzzy systems with constraints on the fuzzy rules. IEEE Trans. Fuzzy Syst. 13, 194\u2013203.                    https:\/\/doi.org\/10.1109\/TFUZZ.2004.840096                                     (2005).\" href=\"\/articles\/s41598-022-20015-y#ref-CR21\" id=\"ref-link-section-d104546385e456\">21<\/a><\/sup>, FN maps the inputs to the outputs by means of connections<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" title=\"Gegov, A., Arabikhan, F. &amp; Petrov, N. Linguistic composition based modelling by fuzzy networks with modular rule bases. Fuzzy Sets Syst. 269, 1\u201329.                    https:\/\/doi.org\/10.1016\/j.fss.2014.06.014                                     (2014).\" href=\"#ref-CR22\" id=\"ref-link-section-d104546385e460\">22<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" title=\"Yaakob, A. M., Gegov, A. &amp; Rahman, S. Selection of alternatives using fuzzy networks with rule base aggregation. Fuzzy Sets Syst. 341, 123\u2013144.                    https:\/\/doi.org\/10.1016\/j.fss.2017.05.027                                     (2018).\" href=\"#ref-CR23\" id=\"ref-link-section-d104546385e460_1\">23<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 24\" title=\"Gegov, A., Sanders, D. &amp; Vatchova, B. Mamdani fuzzy networks with feedforward rule bases for complex systems modelling. J. Intell. Fuzzy Syst. 30, 2623\u20132637.                    https:\/\/doi.org\/10.3233\/IFS-151911                                     (2016).\" href=\"\/articles\/s41598-022-20015-y#ref-CR24\" id=\"ref-link-section-d104546385e463\">24<\/a><\/sup>. The overall number of rules in FN is a linear function of the number of inputs and the number of linguistic terms per input. Compared to Standard Fuzzy System (SFS), the rules in FN are reduced and simplified. The structure in FN is more transparent and interpretable. Arguably, FN is characterized by interpretable innately as a white-box model. The details about FN will be introduced in section&nbsp;\u201c<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"section anchor\" href=\"\/articles\/s41598-022-20015-y#Sec2\">Related methods<\/a>\u201d.<br \/>This paper proposes a hybrid FCM and FN based model for prediction (FCM-FN). FCM-FN is Interpretability-oriented model with Mamdani rule-based structure. The rules are generated by FCM method from data in the first step, and then by expert-system method from expert knowledge in the second step. Afterwards, the rules in both previous steps are connected by FN method. The Interpretable structure of FCM-FN model is identified. As a kind of prediction model, FCM-FN pursues accuracy while preserving interpretable structure. PSO algorithm is utilized to optimize the fuzzy set parameters which is initialized by FCM algorithm in the first step, and by expert system in the second step. For simplicity, the FCM-FN is a kind of multi-input single-output (MISO) type-1 fuzzy system. A multi-input multi-output (MIMO) fuzzy system can be taken as the composition of several MISO fuzzy systems<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 25\" title=\"Gegov, A. Fuzzy Networks for Complex Systems: A Modular Rule Base Approach.                    https:\/\/doi.org\/10.1007\/978-3-642-15600-7                                     (Springer, 2010).\" href=\"\/articles\/s41598-022-20015-y#ref-CR25\" id=\"ref-link-section-d104546385e473\">25<\/a><\/sup>. As for type-2 fuzzy system, the related theories and practices have achieved diverse developments<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 26\" title=\"Mendel, J. M. Uncertain Rule-Based Fuzzy Logic Systems: Introduction and New Directions 2nd edn. (Springer, 2017).\" href=\"\/articles\/s41598-022-20015-y#ref-CR26\" id=\"ref-link-section-d104546385e477\">26<\/a><\/sup>, but the interpretability is rarely included into the type-2 theory so far. This is the reason why the FCM-FN is limited in type-1 structure in this paper.<br \/>The rest of the paper is synthesized as follows: section&nbsp;\u201c<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"section anchor\" href=\"\/articles\/s41598-022-20015-y#Sec2\">Related methods<\/a>\u201d describes the Mamdani\u2019s fuzzy model, FCM method, PSO method and FN theory, which will be used in this paper. In section&nbsp;\u201c<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"section anchor\" href=\"\/articles\/s41598-022-20015-y#Sec15\">The proposed fuzzy model for prediction<\/a>\u201d the structure of the proposed FCM-FN model is analyzed in detail. The effectiveness of the model is illustrated in section&nbsp;\u201c<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"section anchor\" href=\"\/articles\/s41598-022-20015-y#Sec24\">Case study<\/a>\u201d, through applications to various real datasets. The paper ends with the concluding remarks in section&nbsp;\u201c<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"section anchor\" href=\"\/articles\/s41598-022-20015-y#Sec29\">Conclusion<\/a>\u201d, where the proposed approach is summarized, and its main characteristics are identified.<br \/>In this section, the related methods mentioned in the previous section are now further explained. They are divided in five parts: Mamdani-type fuzzy inference, FCM, PSO, PSO based tuning membership functions, and FN.<br \/>Inspired by Zadeh<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 27\" title=\"Zadeh, L. A. Outline of a new approach to the analysis of complex systems and decision processes. IEEE Trans. Syst. Man Cybern. 3, 28\u201344.                    https:\/\/doi.org\/10.1109\/TSMC.1973.5408575                                     (1973).\" href=\"\/articles\/s41598-022-20015-y#ref-CR27\" id=\"ref-link-section-d104546385e511\">27<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 28\" title=\"Zadeh, L. A. Fuzzy sets. Inf. Control 8, 338\u2013353.                    https:\/\/doi.org\/10.1016\/S0019-9958(65)90241-X                                     (1965).\" href=\"\/articles\/s41598-022-20015-y#ref-CR28\" id=\"ref-link-section-d104546385e514\">28<\/a><\/sup>, one of the most interpretable fuzzy models was suggested<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 29\" title=\"Mamdani, E. H. Application of fuzzy logic to approximate reasoning using linguistic synthesis. Comput. IEEE Trans. 100, 1182\u20131191.                    https:\/\/doi.org\/10.1109\/TC.1977.1674779                                     (1977).\" href=\"\/articles\/s41598-022-20015-y#ref-CR29\" id=\"ref-link-section-d104546385e518\">29<\/a><\/sup>, in which Mamdani attempted to control a steam engine and boiler combination by synthesizing a set of linguistic rules from experienced human operators (expert knowledge). The literature on the Mamdani-type fuzzy logic has grown rapidly. One of them was implemented as an universal approximator<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 17\" title=\"Wang, L.-X. &amp; Mendel, J. M. Generating fuzzy rules by learning from examples. IEEE Trans. Syst. Man Cybern. 22, 1414\u20131427.                    https:\/\/doi.org\/10.1109\/21.199466                                     (1992).\" href=\"\/articles\/s41598-022-20015-y#ref-CR17\" id=\"ref-link-section-d104546385e522\">17<\/a><\/sup>. Mamdani-type fuzzy inference includes four steps. In order to explain Mamdani-type fuzzy inference, the example<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 17\" title=\"Wang, L.-X. &amp; Mendel, J. M. Generating fuzzy rules by learning from examples. IEEE Trans. Syst. Man Cybern. 22, 1414\u20131427.                    https:\/\/doi.org\/10.1109\/21.199466                                     (1992).\" href=\"\/articles\/s41598-022-20015-y#ref-CR17\" id=\"ref-link-section-d104546385e526\">17<\/a><\/sup> is employed where the <i>M<\/i> rules are formed as Eq.&nbsp;(<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ1\">1<\/a>).<br \/> where j\u2009=\u2009<i>1,2,\u2026,M, x<\/i><sub><i>i<\/i><\/sub><i>(i<\/i>\u2009=\u2009<i>1,2,\u2026,n)<\/i> are the input variables to the fuzzy system, <i>z<\/i> is the output variable, <span class=\"mathjax-tex\">(A_{i}^{j})<\/span> and <i>B<\/i><sup><i>j<\/i><\/sup> are linguistic terms of the linguistic variables <i>x<\/i><sub><i>i<\/i><\/sub> and <i>z<\/i> in the universes of discourse <i>U<\/i> and <i>R<\/i>, respectively.<br \/>The first step is to take the crisp inputs, and determine the degree to which these inputs belong to each of the appropriate fuzzy sets. Membership functions associating weighted inputs define functional overlaps between inputs, and ultimately determine output responses. Because membership functions are graphical representations of the magnitude of input participations in fuzzy logic, a fuzzy set is defined by its membership functions. Let <i>X<\/i> be a set of items, known as the universe, and its elements are denoted by <i>x<\/i>. And, a fuzzy subset A in <i>X<\/i> is characterized by the membership function <i>\u03bc<\/i><sub><i>A<\/i><\/sub><i>(x)<\/i> which is associated with each element <i>x<\/i> in A and a real number in the interval [0, 1]. The membership function <i>\u03bc<\/i><sub><i>A<\/i><\/sub><i>(x)<\/i> maps each element <i>x<\/i> to a membership value, which represents the level of membership of <i>x<\/i> in A. Different membership functions can be associated with different inputs and outputs. In essence, they are weighting factors for the outcomes of fuzzy rules. Gaussian or triangular shape are two well-known membership functions.<br \/>Gaussian membership function is specified by two parameters as Eq.&nbsp;(<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ2\">2<\/a>).<br \/> where<i> \u03b8<\/i> is the position of the peak relative to the universe, <i>\u03c3<\/i> is the standard deviation.<br \/>Symmetric triangular membership function is also specified by two parameters as Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ3\">3<\/a>) and Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig1\">1<\/a>.<br \/>Symmetric triangular membership function.<br \/>The fuzzified inputs are applied to the rule antecedents. As the fuzzy rule has multiple antecedents, the AND or OR fuzzy operation is used to obtain a single number which represents the result of the antecedent evaluation. In this paper, the AND is applied to evaluate the conjunction of the rule antecedents, which is defined as Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ4\">4<\/a>).<br \/>In other application, if the OR fuzzy operation applies the classical fuzzy operation union, which is defined as Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ5\">5<\/a>).<br \/>Now the antecedent evaluation result can be used to the membership function of the consequent, which is commonly based on two methods: clipping and scaling. Clipping method is to cut the consequent membership function by the truth value of the rule antecedent, and scaling method is to adjust the original membership function of the rule consequent by multiplying all its membership degrees at the level of the antecedent truth. Clipping loses some information, but involving less complex and easier to defuzzify. And scaling preserves the original shape of the fuzzy set. For simplification and interpretation, the method clipping is adopted in this paper.<br \/>The membership functions of all consequents clipped in the previous step are combined into a single fuzzy set.<br \/>The centroid technique is a popular defuzzification method. It finds the centroid point representing the center of gravity (COG) of the aggregated fuzzy set <i>R<\/i>. A reasonable estimate can be obtained by centroid deffuzification method, which is defined as Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ6\">6<\/a>).<br \/> where <span class=\"mathjax-tex\">(overline{z}_{{^{{^{j} }} }})<\/span> is the point in <i>R<\/i> at which <span class=\"mathjax-tex\">(mu_{{B^{j} }} (z))<\/span> achieves its maximum value (usually, we assume that <span class=\"mathjax-tex\">(mu_{{B^{j} }} left( {overline{z}_{{^{{^{j} }} }} } right))<\/span>\u2009=\u20091).<br \/>This method is a universal approximator, i.e. they can approximate any continuous function on a compact set to an arbitrary accuracy<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 17\" title=\"Wang, L.-X. &amp; Mendel, J. M. Generating fuzzy rules by learning from examples. IEEE Trans. Syst. Man Cybern. 22, 1414\u20131427.                    https:\/\/doi.org\/10.1109\/21.199466                                     (1992).\" href=\"\/articles\/s41598-022-20015-y#ref-CR17\" id=\"ref-link-section-d104546385e1754\">17<\/a><\/sup>.<br \/>Clustering is an unsupervised learning method, which assigns a given set of objects into disjoint groups or clusters by membership degrees between 0 and 1. A high degree value represents a high similarity between the object and the group. FCM is a well-known fuzzy clustering algorithms<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 30\" title=\"Full, J. C. B. R. E. W. FCM: The fuzzy c -means clustering algorithm. Comput. Geosci. 10, 191\u2013203.                    https:\/\/doi.org\/10.1016\/0098-3004(84)90020-7                                     (1984).\" href=\"\/articles\/s41598-022-20015-y#ref-CR30\" id=\"ref-link-section-d104546385e1767\">30<\/a><\/sup>. The main motivation of proposing FCM was to address the deficiency in working with overlapping groups shown by the hard cluster algorithm. The FCM method was developed by researches recently<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" title=\"Askari, S., Montazerin, N. &amp; Zarandi, M. F. Generalized possibilistic fuzzy c-means with novel cluster validity indices for clustering noisy data. Appl. Soft Comput. 53, 262\u2013283.                    https:\/\/doi.org\/10.1016\/j.asoc.2016.12.049                                     (2017).\" href=\"#ref-CR31\" id=\"ref-link-section-d104546385e1771\">31<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" title=\"Rubio, E. et al. An extension of the fuzzy possibilistic clustering algorithm using type-2 fuzzy logic techniques. Adv. Fuzzy Syst. 2017, 53.                    https:\/\/doi.org\/10.1155\/2017\/7094046                                     (2017).\" href=\"#ref-CR32\" id=\"ref-link-section-d104546385e1771_1\">32<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" title=\"Mansoureh, N., Mohammad HosseinFazel, Z. &amp; Susan, B. A fuzzy cluster-validity index based on the topology structure and node attribute in complex networks. Expert Syst. Appl. 187, 115913.                    https:\/\/doi.org\/10.1016\/j.eswa.2021.115913                                     (2022).\" href=\"#ref-CR33\" id=\"ref-link-section-d104546385e1771_2\">33<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 34\" title=\"Kaixin, Z., Yaping, D., Zhiyang, J. &amp; Ye, J. General fuzzy C-means clustering algorithm using Minkowski metric. Signal Process. 188, 108161.                    https:\/\/doi.org\/10.1016\/j.sigpro.2021.108161                                     (2021).\" href=\"\/articles\/s41598-022-20015-y#ref-CR34\" id=\"ref-link-section-d104546385e1774\">34<\/a><\/sup>. In order to keep the simplicity and interpretability, the original FCM method is included in FCM-FN model. The process is expressed as follows<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 9\" title=\"Silva-Filho, T. M., Pimentel, B. A., Souza, R. M. C. R. &amp; Oliveira, A. L. I. Hybrid methods for fuzzy clustering based on fuzzy c-means and improved particle swarm optimization. Expert Syst. Appl. 42, 6315\u20136328.                    https:\/\/doi.org\/10.1016\/j.eswa.2015.04.032                                     (2015).\" href=\"\/articles\/s41598-022-20015-y#ref-CR9\" id=\"ref-link-section-d104546385e1778\">9<\/a><\/sup>:<br \/>Let <span class=\"mathjax-tex\">(Omega = { 1, ldots ,k, ldots ,n})<\/span> be a set with <i>n<\/i> objects. Object <i>k<\/i> is the vector of quantitative variables <span class=\"mathjax-tex\">(X_{k} = { x_{1k}, ldots ,x_{jk}, ldots ,x_{pk}})<\/span> described by <i>p<\/i> variables where <span class=\"mathjax-tex\">(x_{jk} in {varvec{R}})<\/span>. Let <span class=\"mathjax-tex\">(Y = { 1, ldots ,i, ldots ,c})<\/span> be a set of <i>c<\/i> prototypes associated to <i>c<\/i> groups, where each prototype <i>i<\/i> is a vector of quantitative variables <span class=\"mathjax-tex\">(Y_{i} = { y_{1i}, ldots ,y_{ji}, ldots ,y_{pi}})<\/span>, where <span class=\"mathjax-tex\">(y_{ji} in {varvec{R}})<\/span>. Let <span class=\"mathjax-tex\">(U = left[ {u_{ik}} right])<\/span> be a <i>c<\/i>\u2009\u00d7\u2009<i>n<\/i> membership degree matrix, where <span class=\"mathjax-tex\">({text{u}}_{ik})<\/span> is the membership degree of object <i>k<\/i> to group <i>i<\/i>, where <span class=\"mathjax-tex\">({text{u}}_{ik} in [0,1])<\/span>. The algorithm works according to minimizing the objective function that is defined as Eq.(7). At the same time, a prototype matrix <span class=\"mathjax-tex\">(Y^{*})<\/span> and a membership degree matrix <span class=\"mathjax-tex\">(U^{*})<\/span> are obtained.<br \/> where <i>m<\/i> is the value of cluster fuzziness, and <span class=\"mathjax-tex\">(d_{ik})<\/span> is the squared Euclidean distance which measures the dissimilarity of the feature vectors between <span class=\"mathjax-tex\">(x_{k})<\/span> and <span class=\"mathjax-tex\">(y_{i})<\/span>. The distance is calculated by Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ8\">8<\/a>).<br \/>Under the minimizing criterion <i>J<\/i>, the prototypes are updated according to Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ9\">9<\/a>), and the membership degrees are updated by Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ10\">10<\/a>) with the restriction <span class=\"mathjax-tex\">(sumlimits_{i = 1}^{c} {u_{ik} = 1})<\/span>.<br \/>As mentioned in section&nbsp;\u201c<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"section anchor\" href=\"\/articles\/s41598-022-20015-y#Sec1\">Introduction<\/a>\u201d, PSO is a typical optimization algorithm that searches for the best solution by modeling the social behavior of bird flocks and fish schools. The population of the birds is named \u201cswarm\u201d, and the members of the population are called \u201cparticles\u201d. Assume that the Search space dimension is D, and the swarm is <i>Sw<\/i>\u2009=\u2009<span class=\"mathjax-tex\">({ X_{1} , ldots ,X_{s} })<\/span>, where <i>s<\/i> is the total number of particles; A particle represents a position in D; the position of the <i>ith<\/i> particle in the search space is denoted as <span class=\"mathjax-tex\">(X_{i} = left( {x_{i1} , ldots ,x_{id} , ldots ,x_{iD} } right),;)<\/span> <i>i<\/i>\u2009=\u2009<i>1,2,\u2026,s<\/i>. <i>pbest<\/i> records the position of <i>ith<\/i> particle\u2019s previous optimal performance, which is denoted as <span class=\"mathjax-tex\">(p_{i} = left( {p_{i1} , ldots ,p_{id} , ldots ,p_{iD} } right))<\/span>. <i>gbest<\/i> is the best position achieved by the swarm, which is expressed as <span class=\"mathjax-tex\">(p_{g} = left( {p_{g1} , ldots ,p_{gd} , ldots ,p_{gD} } right))<\/span>. The velocity of the ith particle is denoted as <span class=\"mathjax-tex\">(v_{i} = left( {v_{i1} , ldots ,v_{id} , ldots ,v_{iD} } right))<\/span>. <i>pbest<\/i> and <i>gbest<\/i> direct a particle\u2019s new velocity and position. <span class=\"mathjax-tex\">(p_{id}^{{^{{}} }} (t))<\/span> is the <i>i<\/i>th particle\u2019s optimal position, and <span class=\"mathjax-tex\">(p_{gd}^{{}} (t))<\/span> is the global best position of the <i>d<\/i>th dimension at instant <i>t<\/i>. The particles fly through the solution space towards better positions, and the process is implemented by Eqs. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ11\">11<\/a>) and (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ12\">12<\/a>).<br \/> where the positive constants <span class=\"mathjax-tex\">(c_{1})<\/span> and <span class=\"mathjax-tex\">(c_{2})<\/span> are acceleration coefficients, <i>r<\/i><sub><i>1<\/i><\/sub> and <i>r<\/i><sub><i>2<\/i><\/sub> are random values in range [0,1]. <span class=\"mathjax-tex\">(x_{id} {text{(t)}})<\/span> is the position and <span class=\"mathjax-tex\">(v_{id} (t))<\/span> is the velocity of <i>i<\/i>th particle in <i>d<\/i>th dimension at instant <i>t<\/i>. The inertia value <span class=\"mathjax-tex\">(omega (t))<\/span> is obtained by Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ13\">13<\/a>).<br \/> where <i>t<\/i> is the current instant and <span class=\"mathjax-tex\">(t_{max })<\/span> is the maximum iteration number. When iteration proceeds, the weighting factor of updating speed will decrease from the maximum factor <span class=\"mathjax-tex\">(omega_{max })<\/span> to the minimum one <span class=\"mathjax-tex\">(omega_{min })<\/span><sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 11\" title=\"Izakian, H. &amp; Abraham, A. Fuzzy C-means and fuzzy swarm for fuzzy clustering problem. Expert Syst. Appl. 38, 1835\u20131838.                    https:\/\/doi.org\/10.1016\/j.eswa.2010.07.112                                     (2011).\" href=\"\/articles\/s41598-022-20015-y#ref-CR11\" id=\"ref-link-section-d104546385e4140\">11<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 15\" title=\"Chen, S., Xu, Z. &amp; Tang, Y. A hybrid clustering algorithm based on fuzzy c-means and improved particle swarm optimization. Arab. J. Sci. Eng. 39, 8875\u20138887.                    https:\/\/doi.org\/10.1007\/s13369-014-1424-9                                     (2014).\" href=\"\/articles\/s41598-022-20015-y#ref-CR15\" id=\"ref-link-section-d104546385e4143\">15<\/a><\/sup>.<br \/>To solve clustering problems, each particle represents a feasible solution to the optimization problem. Let <i>f<\/i> is an object function, and the personal best position of a particle at instant <i>t<\/i> is updated by Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ14\">14<\/a>).<br \/> where <span class=\"mathjax-tex\">(p_{g}^{{}} (t + 1))<\/span> is the global best position at instant <i>t<\/i>+<i>1<\/i> found by anyone of all particles during the previous steps, defined by Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ15\">15<\/a>).<br \/>In the proposed model, there are two kind of membership functions: Gaussian or triangular. PSO algorithm is specified to adjust the shapes of membership functions by tuning the parameters of membership functions.<br \/>The first tuning is aiming towards the parameters of Gaussian membership functions, which are generated by FCM. This process is abbreviated as FCM-PSO. The steps are described by Algorithm FCM-PSO(<i>\u03a9, c<\/i>), which returns matrix <b><i>Y<\/i><\/b> including two kinds of parameters(<i>\u03b8<\/i> and <i>\u03c3<\/i>) in Gaussian membership functions, as shown in Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ2\">2<\/a>).<br \/>The second tuning is aiming towards the parameters of symmetric triangular membership functions, which is specified by experts. This process, similar to FCM-PSO, returns matrix <b><i>M<\/i><\/b> including two kinds of parameters(<i>a<\/i> and \u03b1) in symmetric triangular membership functions, as shown in Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ3\">3<\/a>) and Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig1\">1<\/a>.<br \/>In this section, the details of the novel theory of FN are expressed, which include the basic theory, the basic operations and three important theorems. All the selected details are involved in the proposed model.<br \/>Standard Fuzzy System (SFS) is the most known type of fuzzy system which is with a single rule base. SFS is characterized by the nature of a black-box where the outputs map the inputs directly, and internal connections are out of consideration<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 35\" title=\"Filev, D. Fuzzy modeling of complex systems. Int. J. Approx. Reason. 5, 281\u2013290.                    https:\/\/doi.org\/10.1016\/B978-1-4832-1450-4.50046-8                                     (1991).\" href=\"\/articles\/s41598-022-20015-y#ref-CR35\" id=\"ref-link-section-d104546385e4678\">35<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 36\" title=\"Fern\u00e1ndez Caballero, A. Contribution of fuzziness and uncertainty to modern artificial intelligence. Fuzzy Sets Syst. 160, 129.                    https:\/\/doi.org\/10.1016\/j.fss.2008.05.002                                     (2009).\" href=\"\/articles\/s41598-022-20015-y#ref-CR36\" id=\"ref-link-section-d104546385e4681\">36<\/a><\/sup>. Reflecting the influence of all inputs on the output simultaneously, SFS is quite accurate for output modeling usually. But, when the rules of SFS increasing, the transparency deteriorates, and it is less clear how the inputs affect the output.<br \/>Chained Fuzzy System (CFS), is with multiple rule bases in cascade, and characterized by the nature of white-box where the outputs map the inputs by connected internal variables<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 37\" title=\"Aja-Fernandez, S. &amp; Alberola-Lopez, C. Matrix modeling of hierarchical fuzzy systems. IEEE Trans. Fuzzy Syst. 16, 585\u2013599.                    https:\/\/doi.org\/10.1109\/TFUZZ.2007.905905                                     (2008).\" href=\"\/articles\/s41598-022-20015-y#ref-CR37\" id=\"ref-link-section-d104546385e4688\">37<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 38\" title=\"Hall, L. O. Rule chaining in fuzzy expert systems. IEEE Trans. Fuzzy Syst. 9, 822\u2013828.                    https:\/\/doi.org\/10.1109\/91.971731                                     (2001).\" href=\"\/articles\/s41598-022-20015-y#ref-CR38\" id=\"ref-link-section-d104546385e4691\">38<\/a><\/sup>. CFS is with an arbitrary structure with the form of connected subsystems. It is applied as a detailed presentation of SFS for improving transparency by explicitly accounting all subsystems\u2019 interactions. And the smaller number of inputs improve the efficiency of CFS. However, because of the error accumulation from the multiple Fuzzification-Inference-Defuzzification (FID) sequences, accuracy may be lost. Hierarchical Fuzzy System (HFS), a special kind of CFS, each subsystem in HFS only with one output and two inputs.<br \/>FN, a novel concept with networked rule bases<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 18\" title=\"Wang, X., Gegov, A., Farzad, A., Chen, Y. &amp; Hu, Q. Fuzzy network based framework for software maintainability prediction. Internat. J. Uncertain. Fuzziness Knowl.-Based Syst. 27, 841\u2013862.                    https:\/\/doi.org\/10.1142\/s0218488519500375                                     (2019).\" href=\"\/articles\/s41598-022-20015-y#ref-CR18\" id=\"ref-link-section-d104546385e4698\">18<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 22\" title=\"Gegov, A., Arabikhan, F. &amp; Petrov, N. Linguistic composition based modelling by fuzzy networks with modular rule bases. Fuzzy Sets Syst. 269, 1\u201329.                    https:\/\/doi.org\/10.1016\/j.fss.2014.06.014                                     (2014).\" href=\"\/articles\/s41598-022-20015-y#ref-CR22\" id=\"ref-link-section-d104546385e4701\">22<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 24\" title=\"Gegov, A., Sanders, D. &amp; Vatchova, B. Mamdani fuzzy networks with feedforward rule bases for complex systems modelling. J. Intell. Fuzzy Syst. 30, 2623\u20132637.                    https:\/\/doi.org\/10.3233\/IFS-151911                                     (2016).\" href=\"\/articles\/s41598-022-20015-y#ref-CR24\" id=\"ref-link-section-d104546385e4704\">24<\/a><\/sup>, is characterized by the nature of a white-box where the outputs map the inputs by connections. Arguably, FN is a hybrid of SFS and HFS. On one hand, the structure is similar between FN and HFS because of the explicit presentation of the interactions and subsystems. On the other hand, after the multiple-rule-bases of FN been simplified to a linguistically equivalent single rule base, the operation of FN is similar to SFS. The process of simplification implements the linguistic composition approach, including the vertical merging and horizontal merging of rule bases in FN. The multiple rule base system in HFS is converted into a FN, and then a FN is composed into a single-rule-base system.<br \/>As a hybrid concept, FN obtains the advantages of both accuracy from SFS and transparency from CFS\/HFS. The structure\u2019s transparency is directly related to interpretation of a prediction model, which has turned out to be same important than accuracy and efficiency for complex systems modelling<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 24\" title=\"Gegov, A., Sanders, D. &amp; Vatchova, B. Mamdani fuzzy networks with feedforward rule bases for complex systems modelling. J. Intell. Fuzzy Syst. 30, 2623\u20132637.                    https:\/\/doi.org\/10.3233\/IFS-151911                                     (2016).\" href=\"\/articles\/s41598-022-20015-y#ref-CR24\" id=\"ref-link-section-d104546385e4711\">24<\/a><\/sup>, and it is the core reason that FN is employed in the fuzzy model for prediction in this paper.<br \/>Inputs and outputs of a FN model take linguistic terms. In this paper, the FN model is with only one output(MISO) and the If\u2013then rules are in form of Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ1\">1<\/a>). For compactness, the linguistic terms of the inputs <span class=\"mathjax-tex\">(left( {left{ {A_{1}^{j} , ldots ,A_{n}^{j} } right}} right))<\/span> and the outputs <span class=\"mathjax-tex\">(left( {left{ {B^{j} } right}} right))<\/span> in Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ1\">1<\/a>) are represented by positive integers. For example, {\u2018small\u2019, \u2018average\u2019, \u2018big\u2019} are encoded as positive integers {\u20181\u2019, \u20182\u2019, \u20183\u2019}.<br \/>In order to illustrate the basic operations in FN, a simple example is shown as Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig2\">2<\/a>. A FN with 3 nodes <i>N<\/i><sub><i>11<\/i><\/sub><i>, N<\/i><sub><i>12<\/i><\/sub><i>, I<\/i><sub><i>21<\/i><\/sub> is described by the Boolean matrices given in Table <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"table anchor\" href=\"\/articles\/s41598-022-20015-y#Tab1\">1<\/a>. For simplicity, Each input (output) is only with two linguistic terms, represented by {\u20181\u2019, \u20182\u2019}.&nbsp;More examples have been included in the folder \u201cMathlab functions\u201d in the <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"supplementary material anchor\" href=\"\/articles\/s41598-022-20015-y#MOESM1\">Supplementary File<\/a>.<br \/>Hierarchical structure example for illustrating the basic operations in FN.<br \/>The rules of <i>N<\/i><sub><i>11<\/i><\/sub> are same to the if\u2013then rules given in Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ16\">16<\/a>). The node of <i>N<\/i><sub><i>12<\/i><\/sub> is similar to <i>N<\/i><sub><i>11<\/i><\/sub>.<br \/>The special node of <i>I<\/i><sub><i>21<\/i><\/sub> is an identity node, and the outputs is same to the inputs. The rules of <i>I<\/i><sub><i>21<\/i><\/sub> is described by if\u2013then rules given in Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ17\">17<\/a>).<br \/>Vertical merging is a kind of binary operation that can be applied to a pair of parallel nodes, i.e. nodes locate in the same layer of a FN. In Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig2\">2<\/a>, the nodes <i>N<\/i><sub><i>11<\/i><\/sub> and <i>I<\/i><sub><i>21<\/i><\/sub> are in the same layer. The vertical merging operation is identical with Boolean matrix Kroneker product. This kind of operation merges the operand nodes from the pair into a single product node. In this case, the inputs to the product node represent the union of the inputs to the operand nodes where the outputs from the product node represent the union of the outputs from the operand nodes. The operation of vertical merging can always be applied due to the ability to concatenate the inputs and the outputs of any two parallel nodes<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 24\" title=\"Gegov, A., Sanders, D. &amp; Vatchova, B. Mamdani fuzzy networks with feedforward rule bases for complex systems modelling. J. Intell. Fuzzy Syst. 30, 2623\u20132637.                    https:\/\/doi.org\/10.3233\/IFS-151911                                     (2016).\" href=\"\/articles\/s41598-022-20015-y#ref-CR24\" id=\"ref-link-section-d104546385e6002\">24<\/a><\/sup>. The symbol \u2018+\u2019 denotes the vertical merging operation.<br \/>In Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig3\">3<\/a>, nodes <i>N<\/i><sub><i>11<\/i><\/sub> and <i>I<\/i><sub><i>21<\/i><\/sub> represent a two-node subnetwork of this FN. This two-node FN can be described by the block-scheme and the topological expression in Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ18\">18<\/a>), The vertical merging of the operand nodes <i>N<\/i><sub><i>11<\/i><\/sub> and <i>I<\/i><sub><i>21<\/i><\/sub> results into a single product node <i>N<\/i><sub><i>11<\/i>+<i>21<\/i><\/sub> which represents a simplified image of the two-node FN in the form of a one-node FN. Due to the concatenation of the inputs to the operand nodes as inputs <i>x<\/i><sub><i>1<\/i><\/sub>, <i>x<\/i><sub><i>2<\/i><\/sub> and <i>x<\/i><sub><i>3<\/i><\/sub> to the product node and the concatenation of the outputs from the operand nodes as outputs <i>z<\/i><sub>1<\/sub> <i>and x<\/i><sub><i>3<\/i><\/sub> from the product node. This node <i>N<\/i><sub><i>11<\/i>+<i>21<\/i><\/sub> can be described by the topological expression in Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ19\">19<\/a>) and the Boolean matrix in Table <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"table anchor\" href=\"\/articles\/s41598-022-20015-y#Tab2\">2<\/a>.<br \/>Block-scheme of two nodes\u2019 vertical merging into one node with 3 inputs and 2 outputs.<br \/>Each node in Eqs. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ18\">18<\/a>) and (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ19\">19<\/a>) is placed in a pair of square brackets \u2018[ ]\u2019. The inputs and the outputs for each node are placed in a pair of simple brackets \u2018( )\u2019 right after the node, where the outputs are separated from the inputs by a vertical slash \u2018|\u2019. The topological expressions in the rest of this paper are with the similar structure.<br \/>Horizontal merging is a binary operation that can be applied to a pair of sequential nodes, i.e. nodes located in the same level of a FN. This operation merges the operand nodes from the pair into a single product node. The product node has the same input as the input to the first operand node and the same output as the output from the second operand node whereas the connection does not appear in the product node.<br \/>As described by the block-scheme in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig4\">4<\/a>, the two-node FN can be expressed by the topological Eq.(20). The symbol \u2018*\u2019 implies that the operation of horizontal merging. The horizontal merging of the operand nodes <i>N<\/i><sub><i>11<\/i>+<i>21<\/i><\/sub> and <i>N<\/i><sub><i>12<\/i><\/sub> results into a single product node <i>N<\/i><sub><i>(11<\/i>+<i>21)*12<\/i><\/sub>. As a simplified image of the two-node FN in the form of one-node, <i>N<\/i><sub><i>(11<\/i>+<i>21)*12<\/i><\/sub> is descried by the topological expression in Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ21\">21<\/a>) and the Boolean Matrix in Table <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"table anchor\" href=\"\/articles\/s41598-022-20015-y#Tab3\">3<\/a>.<br \/>Block-scheme of two nodes\u2019 horizontal merging into one node with 3 inputs and 1 outputs.<br \/>The three theorems are the important theoretical base for the proposed model in this paper<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 22\" title=\"Gegov, A., Arabikhan, F. &amp; Petrov, N. Linguistic composition based modelling by fuzzy networks with modular rule bases. Fuzzy Sets Syst. 269, 1\u201329.                    https:\/\/doi.org\/10.1016\/j.fss.2014.06.014                                     (2014).\" href=\"\/articles\/s41598-022-20015-y#ref-CR22\" id=\"ref-link-section-d104546385e7327\">22<\/a><\/sup>.<br \/>Denoted by the symbol \u2018+\u2019, the vertical merging operation is associative according to Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ22\">22<\/a>).<br \/> where for any three operand nodes <i>M, N, O,<\/i> the vertical merging from bottom to top is equal to their vertical merging from top to bottom.<br \/>Denoted by the symbol \u2018<span class=\"stix\">\u2217<\/span>\u2019, the operation of horizontal merging is associative according to Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ23\">23<\/a>).<br \/> where for any three operand nodes <i>M, N, O,<\/i> the vertical merging from right to left is equal to their vertical merging from left to right.<br \/>A HFS with inputs <span class=\"mathjax-tex\">({ x_{1} ,x_{2} , ldots ,x_{{text{m}}} })<\/span>, network nodes <span class=\"mathjax-tex\">({ N_{11} ,N_{12} , ldots ,N_{{text{1,m &#8211; 1}}} })<\/span>, connections <span class=\"mathjax-tex\">({ z_{1} ,z_{2} , ldots ,z_{{text{m &#8211; 2}}} })<\/span> and a output y, as described by the topological expression in Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ24\">24<\/a>)<br \/>can be characterized as a SFS with the same set of <i>m<\/i> inputs, the same single output, a single network node <i>N<\/i>, and no connections as described by the block-schemes in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig5\">5<\/a> and the topological expression in Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ25\">25<\/a>).<br \/> where <span class=\"mathjax-tex\">({ }N = mathop prod limits_{p = 1}^{m &#8211; 1} left( {N_{1p} + mathop sum limits_{q = p + 1}^{m &#8211; 1} I_{pq} } right))<\/span>, node <i>I<\/i><sub><i>qp<\/i><\/sub> is identity node in level <i>p<\/i> and layer <i>q<\/i>.<br \/>Block-scheme of Fuzzy Networks for theorem expression.<br \/>This section describes the proposed FCM-FN model, the process of construction the model includes 4 steps, as shown in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig6\">6<\/a>. The first step aims at selection of input variables, based on the available numerical data and expert knowledge. The second step aims towards generation of knowledge base. The semantic rules and their data bases (DB1) are constructed from expert knowledge, and the clustering rules and its data base (DB2) are generated from numerical data by FCM algorithm. Then, the parameters of DB2 are optimized by the PSO algorithm. In the third step, the two kind of rule bases in second step are combined based on FN method. The forth step aims to optimize the parameters of DB1 also based on PSO algorithm. After the four steps, the proposed model as Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig7\">7<\/a> is converted into a SFS.<br \/>Flow sheet of construction the FCM-FN model.<br \/>FN structure of the proposed model.<br \/>Being converted into a SFS, the proposed FCM-FN model is a MISO system consisting of four major modules: fuzzification, inference engine, defuzzification and knowledge base. The fuzzification module converts the crisp inputs of independent variables to linguistic terms. And these linguistic terms combined with the ones from experts are processed in fuzzy domain by inference engine based on FN. The knowledge base is composed of the rule base, characterizing the control goals and control policy by a set of rules; and of the data base, containing the linguistic term sets and the membership functions defining their semantics. Finally, the processed output is transformed from fuzzy domain (linguistic terms) to crisp domain (numerical number) by defuzzification module<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 18\" title=\"Wang, X., Gegov, A., Farzad, A., Chen, Y. &amp; Hu, Q. Fuzzy network based framework for software maintainability prediction. Internat. J. Uncertain. Fuzziness Knowl.-Based Syst. 27, 841\u2013862.                    https:\/\/doi.org\/10.1142\/s0218488519500375                                     (2019).\" href=\"\/articles\/s41598-022-20015-y#ref-CR18\" id=\"ref-link-section-d104546385e8219\">18<\/a><\/sup>.<br \/>There is large amount of indices to character a product. To reduce the complexity of a model, only those metrics and attributes with significant contribution to the prediction are selected. In practices, experts always are with the knowledge that some metrics are more important than others. Pearson Correlation Coefficient (<span class=\"mathjax-tex\">(r_{p})<\/span>, expressed as Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ26\">26<\/a>)) and Spearman Correlation Coefficient (<span class=\"mathjax-tex\">(r_{r})<\/span>, expressed as Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ27\">27<\/a>)) indicate the strength and direction of a linear relationship between two variables, which are useful to get top relevant input variables.<br \/> where <i>x<\/i> is values in the first set of data, <i>y<\/i> is values in the second set of data, <i>n<\/i> is total number of values.<br \/> where <span class=\"mathjax-tex\">(d_{i}^{2})<\/span> is the difference between the&nbsp;<i>x<\/i>&nbsp;rank and the&nbsp;<i>y<\/i>&nbsp;rank for each pair of data,&nbsp;<span class=\"mathjax-tex\">(sum {d_{i}^{2} })<\/span> is sum of the squared differences between&nbsp;<i>x<\/i>&nbsp;and&nbsp;<i>y<\/i>&nbsp;variable ranks, <i>n<\/i> is&nbsp;sample size.<br \/>As showed in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig7\">7<\/a>, the selected metrics <i>x<\/i><sub><i>i<\/i><\/sub> (<i>i<\/i>\u2009=\u2009<i>1, 2,\u2026, m<\/i>\u2009+\u2009<i>n<\/i>) are input variables. there are <i>m<\/i>\u2009+\u2009<i>n<\/i> inputs divided into two sets. SET1 is composed of numerical inputs <i>x<\/i><sub><i>i<\/i><\/sub> (<i>i<\/i>\u2009=\u2009<i>1, 2,\u2026, m<\/i>) which are from the variables with numerical data, and SET2 is composed of semantic inputs <i>x<\/i><sub><i>i<\/i><\/sub> (<i>i<\/i>\u2009=\u2009<i>m<\/i>\u2009+\u2009<i>1, m<\/i>\u2009+\u2009<i>2, \u2026, m<\/i>\u2009+\u2009<i>n<\/i>) from the variables only with the expert\u2019s linguistic information. Combined with the numerical data of output <i>z<\/i><sub><i>1<\/i><\/sub> (here <i>z<\/i><sub><i>1<\/i><\/sub> is the prediction variable), the numerical data are transformed into a set of input\u2013output data pairs.<br \/>The process of generation and optimization of knowledge base includes three parts: generation of rule base, generation of data base and parameters optimization of Gaussian membership functions.<br \/>Given a set of input\u2013output data pairs as Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ28\">28<\/a>), the sampling data is preprocessed, including deleting the duplicated data pairs and min\u2013max normalization according to Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ29\">29<\/a>).<br \/> where <span class=\"mathjax-tex\">(x^{(p)} in {text{R}}^{m} ;and;y^{(p)} in {text{R}})<\/span>.<br \/> where <span class=\"mathjax-tex\">(x^{prime})<\/span> is the min\u2013max normalized set, <i>x<\/i> is a set of the observed values present in <i>x<\/i>, <span class=\"mathjax-tex\">(x_{min })<\/span> is the minimum values in <i>x,<\/i> <span class=\"mathjax-tex\">(x_{max })<\/span> is the maximum values in <i>x.<\/i><br \/>The basic problem is to extract rules which describe how the output variable <span class=\"mathjax-tex\">(y in {text{R}})<\/span> is influenced by the <i>m<\/i> input variables <span class=\"mathjax-tex\">(x = (x_{1} , ldots ,x_{m} )^{T} in R^{m})<\/span> based on Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ28\">28<\/a>). FCM is a powerful unsupervised learning technique to extract rules of Mamdani structure as Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ1\">1<\/a>), which divides sampling data into several clusters based on their similarity<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 5\" title=\"Dick, S., Meeks, A., Last, M., Bunke, H. &amp; Kandel, A. Data mining in software metrics databases. Fuzzy Sets Syst. 145, 81\u2013110.                    https:\/\/doi.org\/10.1016\/j.fss.2003.10.006                                     (2004).\" href=\"\/articles\/s41598-022-20015-y#ref-CR5\" id=\"ref-link-section-d104546385e9164\">5<\/a><\/sup>. In the proposed model, <i>K<\/i> is the number of clusters and specified to be a positive odd integer. After the sampling data is clustered by FCM algorithm, the input space of each numerical input is divided into <i>K<\/i> sections, and <i>K<\/i> Mamdani rules is generated. The rule base of Node <i>N<\/i><sub><i>11<\/i><\/sub> in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig7\">7<\/a> is composed by these <i>K<\/i> rules.<br \/>The semantic inputs in SET2 and outputs are Subjective Product Appraisals (SPA). For simplicity and interpretability, the input space of these semantic inputs and outputs are also divided into <i>K<\/i> sections. For example, if <i>K<\/i>\u2009=\u20093, the semantic choice scale is {Low, Nominal, High} presented by the positive integers 1\u20133. These rule bases are derived from the knowledge of experts and engineers. It is worth noting that the quality of the rule bases affects the output to some degree. The rule base for Nodes <i>N<\/i><sub><i>1n<\/i><\/sub> and <i>N<\/i><sub><i>2j<\/i><\/sub> (<i>j<\/i>\u2009=\u2009<i>1,\u2026,n\u22121<\/i>), which are associated with these semantic inputs, is Mamdani type as Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ1\">1<\/a>), and the number of rules is not over <i>K<\/i><sup>2<\/sup> because of only two inputs of each node.<br \/>The identity nodes <i>I<\/i><sub><i>ij<\/i><\/sub> (<i>i<\/i>\u2009=\u2009<i>1,\u2026,n; j<\/i>\u2009=\u2009<i>1,\u2026,n\u22121<\/i>) are with the rule base of Mamdani type, too. And the number of rules is <i>K<\/i> because of only one input of each identity node.<br \/>As shown in Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ1\">1<\/a>), each fuzzy rule has multiple antecedents. The AND fuzzy operation intersection is implemented, which is defined as Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ4\">4<\/a>). The reason is that all the input variables simultaneously affect the output, and rules in antecedent must be met simultaneously in order for consequent to occur. For prediction problems, only \u201cand\u201d rules are required since the antecedents are different components of a single input vector.<br \/>The data base is associated with the type of membership functions. The process of assignment membership functions can be intuitive or based on some algorithmic operations. Six straightforward methods are described in the literature to assign membership functions to fuzzy variables. Based on the judgment about the probability density functions from researchers and ourselves, the shape of membership function of input variables of node <i>N<\/i><sub><i>11<\/i><\/sub> is Gaussian curve, as defined in Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ2\">2<\/a>). The position of the peak relative to the universe and the standard deviation are two parameters of a Gaussian membership function. The parameters are determined by FCM algorithm based on sampling data. In this paper, the result of FCM algorithm is named as FCM model, in which the knowledge base is with parameters of Gaussian membership functions.<br \/>For these semantic inputs, a natural membership function that readily comes to mind is symmetric triangular membership functions defined as Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ3\">3<\/a>). Two main reasons motivate the choice: one is their optimal interface design and the other is its semantic integrity.<br \/>In this step, the optimization is aimed towards the data base of the FCM model and based on the sampling data as Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ3\">3<\/a>). MATLAB-Fuzzy Logic Tool Box (<b>genfis<\/b>) is utilized to generate the FIS of FCM model. To improve the precision as well as reduce the loss of the interpretability, the optimization is only aiming towards parameters (<i>\u03b8<\/i> and \u03c3) of inputs and the output. The structure of this FIS is unchanged. As mentioned in section&nbsp;\u201c<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"section anchor\" href=\"\/articles\/s41598-022-20015-y#Sec9\">PSO<\/a>\u201d, as a global optimization method, PSO algorithm is suitable for initial training to tune the parameters. In the proposed model, the MATLAB function \u201c<b>tunefis<\/b>\u201d is employed with the specified tuning algorithm name as \u201c<b>particleswarm<\/b>\u201d. The optimized model is named as FCM\u2013PSO in this paper. It works by running the FCM method until it reaches its stopping criterion. Then it runs the PSO algorithm to try to achieve a better solution. In the experiments presented in section \u201c<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"section anchor\" href=\"\/articles\/s41598-022-20015-y#Sec24\">Case study<\/a>\u201d, the following stopping criteria were used:<br \/>FCM: when there is a variation less than or equal to 1e\u22128 on Minimum improvement in objective function (MinImprovement for short).<br \/>PSO: when relative change in the best objective function value is less than 1e-8 on FunctionTolerance (a method option of \u201c<b>tunefis<\/b>\u201d) or it reaches maximum number of iterations (MaxNumIteration for short). The experiments explored 11 different MaxNumIterations in the set {10, 20, 30, 40, 50, 100, 200, 300, 400, 1000, 2000}.<br \/>The semantic inputs are organized into HFS structure as shown in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig7\">7<\/a>. Two input variables (<i>x<\/i><sub><i>m<\/i>+<i>1<\/i><\/sub> and <i>x<\/i><sub><i>m<\/i>+<i>2<\/i><\/sub>) are organized into <i>N<\/i><sub><i>21<\/i><\/sub>, and other input variables are added one by one after converted by the identify node. And <i>N<\/i><sub><i>2,n\u22121<\/i><\/sub> combines all <i>n<\/i> semantic inputs. Identify nodes <i>I<\/i><sub><i>1j<\/i><\/sub> (<i>j<\/i>\u2009=\u2009<i>2,\u2026,n\u22121<\/i>) are employed to combine <i>N<\/i><sub><i>2,n\u22121<\/i><\/sub> with <i>N<\/i><sub><i>11<\/i><\/sub>. <i>N<\/i><sub><i>1n<\/i><\/sub> is the final node which combine numerical inputs with semantic inputs.<br \/>To combine these FN nodes into a SFS, the rule bases of all nodes are merged vertically and horizontally<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 39\" title=\"Gegov, A., Petrov, N. &amp; Vatchova, B. In Intelligent Systems (IS), 2010 5th IEEE International Conference 197\u2013202 (IEEE, 2010).\" href=\"\/articles\/s41598-022-20015-y#ref-CR39\" id=\"ref-link-section-d104546385e9399\">39<\/a><\/sup>. The rule bases of <i>N<\/i><sub><i>11<\/i><\/sub><i>, N<\/i><sub><i>21<\/i><\/sub><i>, I<\/i><sub><i>31<\/i><\/sub><i>,<\/i> \u2026 and <i>I<\/i><sub><i>n,1<\/i><\/sub> are merged firstly, the result named rule base <i>V<\/i><sub><i>1<\/i><\/sub>, and in the same method, rule base <i>V<\/i><sub><i>2<\/i><\/sub> is from <i>I<\/i><sub><i>12<\/i><\/sub><i>, N<\/i><sub><i>22<\/i><\/sub><i>,<\/i> \u2026 and <i>I<\/i><sub><i>n,2<\/i><\/sub>, rule base <i>V<\/i><sub><i>n\u22121<\/i><\/sub> is from the last vertical merging of <i>I<\/i><sub><i>1,n\u22121<\/i><\/sub> and <i>N<\/i><sub><i>2,n\u22121<\/i><\/sub>. And then, all the rule base of <i>V<\/i><sub><i>i<\/i><\/sub><i> (0<\/i>\u2009&lt;\u2009<i>i<\/i>\u2009\u2264\u2009<i>n\u22121)<\/i> are merged horizontally. Till now, the single rule base RB of the proposed model is generated, as topological expression in Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ30\">30<\/a>).<br \/> where \u201c\u2009+\u2009\u201d means vertical merging operation and \u201c\u2009<span class=\"stix\">\u2217<\/span>\u2009\u201d means horizontal merging operation.<br \/>After the multiple rule bases are merged into a linguistically equivalent single rule base, the FN model is converted to a kind of SFS model with <i>m<\/i>\u2009+\u2009<i>n<\/i> inputs, a output and a knowledge base RB, as shown in Fig. <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig8\">8<\/a>.<br \/>SFS structure after knowledge base combination based on FN method.<br \/>In this step, the optimization is implemented to the data base of <i>N<\/i><sub><i>2j<\/i><\/sub> (<i>j<\/i>\u2009=\u2009<i>1,\u2026,n\u22121<\/i>) in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig7\">7<\/a>. The initial values of parameters (<i>a<\/i> and <i>\u03b1<\/i>) of symmetric triangular membership functions, which was specified by experts, are tuned by PSO algorithm.<br \/>To implement the tuning, SPAs of <i>N<\/i><sub><i>2j<\/i><\/sub> should be transformed to numerical data. For example, the Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ31\">31<\/a>) can finish this transform. The semantic appraisal for each semantic input is from the experts\u2019 judgments.<br \/> where <i>i<\/i> is the <i>i<\/i>th item in the linguistic term set. <i>K<\/i> is the number of items of the linguistic term set.<br \/>For example, the linguistic term set {small, average, big} have 3 items, in which the item \u201caverage\u201d can be converted to 0.5, as shown in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig9\">9<\/a>A. And {very small, small, average, big, very big} have 5 items, and the item \u201cbig\u201d is same to 0.75, as shown in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig9\">9<\/a>B. What\u2019s more, the experts can appraise a semantic input by using a real number in the interval [0, 1] directly with the help of membership functions.<br \/>Symmetric triangular membership function with 3 linguistic terms and 5 linguistic terms respectively.<br \/>Combining <span class=\"mathjax-tex\">({text{D}}_{SPAi})<\/span> of <i>x<\/i><sub><i>m<\/i>+<i>1<\/i><\/sub><i> to x<\/i><sub><i>m<\/i>+<i>n<\/i><\/sub> with data pairs(<i>x<\/i><sub><i>1<\/i><\/sub> to <i>x<\/i><sub><i>m<\/i><\/sub><i>,<\/i> and <i>y<\/i>) produces new data pairs for FN. Two different type of input data are included in the new data pairs. One is from numerical inputs and this kind of inputs generally is quantifiable. For example, Lines of a paragraph of software code can be considered as numerical input. The other is from qualitative inputs. For example, programming language complexity is always appraised by semantic terms.<br \/>The new data pairs for FN is utilized to tune the parameters of triangular membership functions. The process of optimization is similar to section&nbsp;\u201c<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"section anchor\" href=\"\/articles\/s41598-022-20015-y#Sec20\">Parameters optimization of Gaussian membership functions<\/a>\u201d, and with the same stopping criteria.<br \/>After being optimized in this step, the knowledge base RB of the proposed model is renewed and ready for prediction.<br \/>The proposed model with optimized knowledge base has been converted into a SFS model as shown in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig8\">8<\/a> in the previous steps. Given a data pairs of <i>m<\/i> numerical inputs and <i>n<\/i> semantic inputs, the prediction value is yielded according to the FCM-FN model.<br \/>In this section, a simple FCM-FN model was applied to 6 real world datasets to evaluate the performances. The details are described in 4 parts: the simple FCM-FN model, the performance metrics, the 6 datasets, and the results of case study. The model was implemented on a computer with an Intel Core i7\u20139750, 2.60&nbsp;GHz processor and 16&nbsp;GB RAM, running a Windows 10(64-bit) operating system and MATLAB R2019a.<br \/>Prior to building the rules, the number of linguistic variables is limited to a considerable size to avoid excessive rule explosion and deteriorated interpretation. This is because potentially, every combination of variables could require a distinct rule. Due to this, the linguistic terms used in the performance evaluation are limited to 3 which are {poor, fair and good}, or {low, moderate and high}<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 40\" title=\"Olugu, E. U. &amp; Wong, K. Y. An expert fuzzy rule-based system for closed-loop supply chain performance assessment in the automotive industry. Expert Syst. Appl. 39, 375\u2013384.                    https:\/\/doi.org\/10.1016\/j.eswa.2011.07.026                                     (2012).\" href=\"\/articles\/s41598-022-20015-y#ref-CR40\" id=\"ref-link-section-d104546385e10157\">40<\/a><\/sup>.<br \/>In order to show the interpretation and accuracy of the model, a simple FCM-FN model with only 3 numerical inputs and 3 semantic inputs is set up, as shown in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig10\">10<\/a>. And 6 real world datasets are selected, which are descripted in section&nbsp;\u201c<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"section anchor\" href=\"\/articles\/s41598-022-20015-y#Sec27\">Datasets<\/a>\u201d. For selection of the input variables <i>x<\/i><sub><i>1<\/i><\/sub> to <i>x<\/i><sub><i>3<\/i><\/sub> of each dataset, Pearson&#8217;s and Spearman\u2019s Correlation Coefficients are utilized to select top 3 relevant input variables according to the strength and direction of a linear relationship between each input and the output. The definitions of inputs <i>x<\/i><sub><i>4<\/i><\/sub> to <i>x<\/i><sub><i>6<\/i><\/sub> depend on the needs of prediction. The data pairs of all 6 inputs and a output is nondimensionalized according to Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ4\">4<\/a>).<br \/>Simple FCM-FN model with 3 numerical inputs and 3 sematic inputs.<br \/>In this simple FCM-FN model, the number of linguistic sets is specified to 3(<i>K<\/i>\u2009=\u20093). It means that the clusters in FCM is 3, <i>x<\/i><sub><i>1<\/i><\/sub> to <i>x<\/i><sub><i>3<\/i><\/sub>, <i>z<\/i><sub><i>1<\/i><\/sub>, <i>y<\/i> are with 3 Gaussian membership functions and <i>x<\/i><sub><i>4<\/i><\/sub> to <i>x<\/i><sub><i>6<\/i><\/sub>, <i>z<\/i><sub><i>2<\/i><\/sub>, <i>z<\/i><sub><i>3<\/i><\/sub> are with 3 symmetric triangular membership functions, separately.<br \/>Each node of <i>N<\/i><sub><i>21<\/i><\/sub>, <i>N<\/i><sub><i>22<\/i><\/sub>, <i>N<\/i><sub><i>13<\/i><\/sub> is with 9 (=\u20093<sup>2<\/sup>) rules. For example, in order to predict the software\u2019s quality, the input variable <i>x<\/i><sub><i>4<\/i><\/sub> is specified as extent of the supporting document set, <i>x<\/i><sub><i>5<\/i><\/sub> is comprehensibility of the supporting document set and <i>x<\/i><sub><i>6<\/i><\/sub> is programming language complexity. They are organized into hierarchical structure, as shown in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig10\">10<\/a>. The structure was not easy to make sense because <i>z<\/i><sub><i>2<\/i><\/sub> and z<sub>3<\/sub> should be with exact connotation. Here, <i>z<\/i><sub><i>2<\/i><\/sub> is the subjective complexity appraisal, and z<sub>3<\/sub> is subjective supporting environment appraisal. As shown in Table <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"table anchor\" href=\"\/articles\/s41598-022-20015-y#Tab4\">4<\/a>, the rule bases of <i>N<\/i><sub><i>21<\/i><\/sub><i>, N<\/i><sub><i>22<\/i><\/sub> and <i>N<\/i><sub><i>13<\/i><\/sub> are presented by Boolean matrices and derived by the knowledge of software engineer experts.<br \/>The knowledge base of <i>N<\/i><sub><i>11<\/i><\/sub> is generated by FCM algorithm, and the rule base includes 3 rules, and each identification node is also with 3 rules. Being merged based on FN method, the rule base of the simple FCM-FN model is with 81 rules.<br \/>As opposed to most existing approaches where the focus is to improve accuracy, the FCM-FN method focuses to maintain interpretation, transparency and accuracy by means of the modular rule bases that reflect the subsystems of the modeled system. There are three performance indicators to show the quality of the associated models. They are called Transparency Index (<i>TI<\/i>)<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 22\" title=\"Gegov, A., Arabikhan, F. &amp; Petrov, N. Linguistic composition based modelling by fuzzy networks with modular rule bases. Fuzzy Sets Syst. 269, 1\u201329.                    https:\/\/doi.org\/10.1016\/j.fss.2014.06.014                                     (2014).\" href=\"\/articles\/s41598-022-20015-y#ref-CR22\" id=\"ref-link-section-d104546385e11370\">22<\/a><\/sup>, Interpretation Index(<i>PI<\/i>) and Accuracy Index (<i>AI<\/i>).<br \/>The first performance indicator <i>TI<\/i> reflects the transparency according to the extent of its opaqueness from the inside of a model, as shown by Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ32\">32<\/a>).<br \/> where <i>t<\/i> is the total number of inputs, 1 is refer to the only one output, <i>p<\/i> is the number of non-identity nodes and <i>q<\/i> is the number of non-identity connections. A lower <i>TI<\/i> implies better transparency.<br \/>For example, in the simple FCM-FN model as shown in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig10\">10<\/a>, <i>t<\/i>\u2009=\u20096, <i>p<\/i>\u2009=\u20094, <i>q<\/i>\u2009=\u20093,and <i>TI<\/i>\u2009=\u20091.<br \/>The second performance indicator <i>PI<\/i> reflects the interpretability of a model. It\u2019s obvious that transparency is helpful to interpretability and more parameters impact the interpretability. Thus, we put forward a modified <i>TI<\/i> to reflects the interpretability of a FCM-FN model, as shown by Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ33\">33<\/a>).<br \/> where <i>r<\/i> is the total number of parameters, <i>s<\/i> is the total number of optimized parameters. A lower <i>PI<\/i> implies better interpretability.<br \/>For example, in the model showed in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig10\">10<\/a>, <i>t<\/i>\u2009=\u20096, <i>p<\/i>\u2009=\u20094, <i>q<\/i>\u2009=\u20093, <i>r<\/i>\u2009=\u200920, <i>s<\/i>\u2009=\u200914, and <i>PI<\/i>\u2009=\u20091.646.<br \/>The third performance indicator <i>AI<\/i> reflects the accuracy of a model by means of Symmetric Mean Absolute Percentage Error (<i>SMAPE<\/i>) as shown by Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ35\">35<\/a>) and Goodness-of-Fit (<i>R<\/i><sup><i>2<\/i><\/sup>) as shown by Eq. (<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"equation anchor\" href=\"\/articles\/s41598-022-20015-y#Equ9\">9<\/a>).<br \/> where <i>n<\/i> is the total number of observations compared, <i>actual<\/i> is the <i>i<\/i>th value of the test dataset estimated values vector, and <i>prediction<\/i> is the <i>i<\/i>th value of output vector of a model, <i>mean<\/i> is the mean value of <i>actual<\/i>. Lower <i>SMAPE<\/i> or higher <i>R<\/i><sup>2<\/sup> implies better accuracy.<br \/>Six well-known real-world datasets are loaded to evaluate the performances of the FCM-FN model compared with FCM model and FCM-PSO model&nbsp;(See <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"supplementary material anchor\" href=\"\/articles\/s41598-022-20015-y#MOESM1\">Supplementary File<\/a> &#8211; raw dataset).<br \/>Table <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"table anchor\" href=\"\/articles\/s41598-022-20015-y#Tab5\">5<\/a> is a summary of the 6 datasets. These datasets are cases proposed in previous studies in various fields including social, material, medical and software. Based on Pearson&#8217;s and Spearman\u2019s Correlation Coefficients, only top 3 attributes and the prediction target of each database are selected as simplified benchmark dataset to evaluate the performance. The number of instances is the rows of original databases, and the number of unique instances is the rows of benchmark dataset after deleted the duplicated instances.<br \/>White wine dataset and red wine dataset were obtained from the same source and with the same structure<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 41\" title=\"Cortez, P., Cerdeira, A., Almeida, F., Matos, T. &amp; Reis, J. Modeling wine preferences by data mining from physicochemical properties. Decis. Support Syst. 47, 547\u2013553.                    https:\/\/doi.org\/10.1016\/j.dss.2009.05.016                                     (2009).\" href=\"\/articles\/s41598-022-20015-y#ref-CR41\" id=\"ref-link-section-d104546385e12246\">41<\/a><\/sup>. The two datasets were related to white and red variants of the Portuguese &#8220;Vinho Verde&#8221; wine. Both datasets contain chemical analysis of wine derived from three different cultivars grown in the same region. Eleven physicochemical attributes of wine were included: 1\u2014fixed acidity, 2\u2014volatile acidity, 3\u2014citric acid, 4\u2014residual sugar, 5\u2014chlorides, 6\u2014free sulfur dioxide, 7\u2014total sulfur dioxide, 8\u2014density, 9\u2014pH, 10\u2014sulphates, 11\u2014alcohol. The prediction target was wine quality, which was graded between 0 (very bad) and 10 (very excellent) by wine experts.<br \/>Concrete dataset recorded 1030 instances and 9 attributes. The prediction target was concrete compressive strength, and the 8 input variables including 1\u2014cement, 2\u2014blast furnace slag, 3\u2014fly ash, 4\u2014water, 5\u2014super plasticizer, 6\u2014coarse aggregate, 7\u2014fine aggregate and 8-age. The concrete compressive strength was a highly nonlinear function of age and 7 ingredients. And the actual concrete compressive strength (MPa) for a given mixture under a specific age (days) was determined from laboratory<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 42\" title=\"Yeh, I.-C. Modeling of strength of high-performance concrete using artificial neural networks. Cem. Concr. Res. 28, 1797\u20131808.                    https:\/\/doi.org\/10.1016\/s0008-8846(98)00165-3                                     (1998).\" href=\"\/articles\/s41598-022-20015-y#ref-CR42\" id=\"ref-link-section-d104546385e12253\">42<\/a><\/sup>.<br \/>Boston dataset recorded 506 house prices. The dataset was taken from the StatLib library which was maintained at Carnegie Mellon University. The original dataset contained 14 attributes. In order to avoid racism, only 13 attributes was contained in the newly version dataset: 1\u2014CRIM (per capita crime rate by town), 2\u2014ZN (proportion of residential land zoned for lots over 25,000 sq.ft.), 3\u2014INDUS (proportion of non-retail business acres per town), 4\u2014CHAS (Charles River dummy variable), 5\u2014NOX(nitric oxides concentration), 6\u2014RM (average number of rooms per dwelling), 7\u2014AGE (proportion of owner-occupied units built prior to 1940), 8\u2014DIS (weighted distances to five Boston employment centres), 9\u2014RAD (index of accessibility to radial highways), 10\u2014TAX (full-value property-tax rate per $10,000), 11\u2014PTRATIO (pupil-teacher ratio by town), 12\u2014LSTAT (lower status of the population), 13\u2014MEDV (Median value of owner-occupied homes in $1000&#8217;s). And MEDV was usually the prediction target.<br \/>Diabetes recorded 442 patients health conditions by 11 attributes: 1\u2014age, 2\u2014sex, 3\u2014bmi (Body mass index), 4\u2014bp (Average Blood Pressure), 5\u2014s1, 6\u2014s2, 7\u2014s3, 8\u2014s4, 9\u2014s5, 10\u2014s6, 11\u2014outcome. And s1\u2013s6 were 6 serum test data one year later. The prediction target was outcome (quantitative indicator of diabetes one year later).<br \/>QUES database was selected from popular object-oriented maintainability datasets which were published by Li and Henry<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 43\" title=\"Li, W. &amp; Henry, S. Object-oriented metrics that predict maintainability. J. Syst. Softw. 23, 111\u2013122.                    https:\/\/doi.org\/10.1016\/0164-1212(93)90077-B                                     (1993).\" href=\"\/articles\/s41598-022-20015-y#ref-CR43\" id=\"ref-link-section-d104546385e12267\">43<\/a><\/sup>. Ten attributes were included as input variables: 1\u2014WMC (Weighted method per class), 2\u2014DIT (Depth of the inheritance tree), 3\u2014NOC (Number of children), 4\u2014RFC (Response for class), 5\u2014LCOM (Lack of cohesion of methods), 6\u2014MPC (Message-passing coupling), 7\u2014DAC (Data abstraction coupling), 8\u2014NOM (Number of methods), 9\u2014SIZE2 (Number of properties), 10\u2014SIZE1 (Lines of code). The prediction target was CHANGE (Number of lines changed in the class), which recorded the number of changed lines in the code during a three-year maintenance period. The values of 11 attributes in QUES was collected from 71 classes in a software system.<br \/>In this section, the results of the proposed model on the real-world datasets are presented. The comparison is implemented among FCM model, FCM-PSO model and FCM-FN model. FCM model is only related to the node <i>N<\/i><sub><i>11<\/i><\/sub>. After the parameters being tuned, FCM model is transformed into FCM-PSO. And after implementing the FN method and PSO tuning, FCM-PSO model is transformed into FCM-FN.<br \/>Table <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"table anchor\" href=\"\/articles\/s41598-022-20015-y#Tab6\">6<\/a> summarized the performance <i>TI<\/i> and <i>PI<\/i> of the three models: FCM, FCM-PSO, FCM-FN. <i>TI<\/i> is related to the structure of models directly. The FCM and FCM-PSO have the same structure and <i>TI<\/i> (=\u20094). By contrast, the FCM-FN model is with non-identity nodes and connections and lower <i>TI<\/i> (=\u20091). <i>PI<\/i> is related to both the structure and parameters. Though FCM and FCM-PSO have the same structure, but the latter tunes the parameters and then impact <i>PI,<\/i> which is added from 5.021 to 5.660. Due to lower <i>TI<\/i>, the <i>PI<\/i> of FCM-FN is 1.646 and far lower than the other two models.<br \/>Table <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"table anchor\" href=\"\/articles\/s41598-022-20015-y#Tab7\">7<\/a> are the results of performance <i>SMAPE<\/i> of the three models utilizing the 6 benchmark datasets. The <i>AI<\/i> of each model is appraised with the process of ten-fold cross-validation. Each dataset is randomly shuffled and then divided into 10 partitions. 9 partitions for training and 1 partition for testing. The value of <i>AI<\/i> is the mean value after running each model 10 times, in format of percentage. In Table <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"table anchor\" href=\"\/articles\/s41598-022-20015-y#Tab7\">7<\/a>, the \u2018%\u2019 is omitted in order to save space, and the values of best accuracy among the three models are bolded.<br \/>In order to explore the effects of different iterations of PSO, 11 different values from 10 to 2000 are set separately as the MaxNumIteration parameter of PSO. The values of <i>SMAPE<\/i> after 11 different iterations are shown in Table <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"table anchor\" href=\"\/articles\/s41598-022-20015-y#Tab7\">7<\/a>. It\u2019s obvious that the accuracy of FN-PSO and FCM-PSO is far better than FCM model, and FN-PSO better than FCM-PSO in most instances.<br \/>The means of 11 <i>SMAPE<\/i> values of each model is shown in Table <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"table anchor\" href=\"\/articles\/s41598-022-20015-y#Tab8\">8<\/a>, as well as the improvements of FCM-FN compared with FCM and FCM-PSO respectively. The accuracy of the FCM-FN model is improved from 5.012% to 30.784% compared with FCM, and from 0.204% to 1.201% compared with FCM-PSO. The average <i>SMAPE<\/i> of each dataset shows that FCM-FN models marginally outperforms FCM-PSO models. This observation is not surprising since FCM-FN model derived most of its initial semantic inputs from the experts\u2019 knowledge.<br \/>In general, increasing number of iterations in parameter-tuning phase, the <i>SMAPE<\/i> and <i>R<\/i><sup>2<\/sup> is increasing accordingly as shown in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig11\">11<\/a>A\u2013C and Table <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"table anchor\" href=\"\/articles\/s41598-022-20015-y#Tab7\">7<\/a>. But, as shown in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig11\">11<\/a>D\u2013F, the <i>R<\/i><sup><i>2<\/i><\/sup> is not always increasing along with the iterations. What\u2019s more, as shown in Fig.&nbsp;<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"\/articles\/s41598-022-20015-y#Fig11\">11<\/a>F, the <i>R<\/i><sup><i>2<\/i><\/sup> fluctuates sharply because there are only 71 instances. As shown in Table <a data-track=\"click\" data-track-label=\"link\" data-track-action=\"table anchor\" href=\"\/articles\/s41598-022-20015-y#Tab7\">7<\/a>, the PSO algorithm always converged before reaching the aforementioned MaxNumIteration. For example, The FN-PSO and FCM-PSO reached their best accuracy when the benchmark dataset was red wine and MaxNumIteration was 200, and more MaxNumIteration was not yielded higher accuracy because the FCM-PSO converged no more than 257 iterations and FN-PSO no more than 51 iterations. And doing so increased the duration of the optimization process and also increased validation error due to overtuned system parameters with the training data. Thus, the following accuracy values fluctuated around the best one.<br \/>The <i>R<\/i><sup>2<\/sup> of the three models after different iterations\u2019 optimization.<br \/>To improve the performances of prediction model both in accuracy and transparency with well tradeoff interpretability, this paper proposed a FCM-FN model which combines FCM, FN and PSO methods. The main motivation of utilizing FN in prediction model is that interpretability is the dominant feature of a fuzzy model in security-oriented field and other special fields. As noted earlier, FN method has been characterized by the nature of a white-box where the output variable map the input metrics by connections, and the implied information in the FN model has been shown totally. As to the performance of <i>TI<\/i> and <i>PI<\/i>, FCM-FN model is overwhelming compared to FCM model or FCM-PSO model.<br \/>Based on FCM method, the numerical data is clustered and then denoted by rules with simple and interpretable structure. Optimized with PSO method, the accuracy of FCM-FN model has been improved significantly (over 5%).<br \/>In conclusion, the FCM-FN model has more transparency, interpretability and accuracy than FCM model or FCM-PSO model. Moreover, FCM-FN model is a special type of SFS, because the numerical data and linguistic appraisals can be combined into the prediction model at the same time. Although this strategy is not perfect, it tends to catch more subjective appraisals which have shown their powerful efficacy in previous researches<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 44\" title=\"Van den Broeke, M., De Baets, S., Vereecke, A., Baecke, P. &amp; Vanderheyden, K. Judgmental forecast adjustments over different time horizons. Omega 87, 34\u201345.                    https:\/\/doi.org\/10.1016\/j.omega.2018.09.008                                     (2019).\" href=\"\/articles\/s41598-022-20015-y#ref-CR44\" id=\"ref-link-section-d104546385e13984\">44<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 45\" title=\"Arvan, M., Fahimnia, B., Reisi, M. &amp; Siemsen, E. Integrating human judgement into quantitative forecasting methods: A review. Omega 86, 237\u2013252.                    https:\/\/doi.org\/10.1016\/j.omega.2018.07.012                                     (2019).\" href=\"\/articles\/s41598-022-20015-y#ref-CR45\" id=\"ref-link-section-d104546385e13987\">45<\/a><\/sup>.<br \/>In this paper only 3 subjective appraisal variables are specified as the example, and the results indicate that it is possible to find qualitative rules. We do not claim that the rules between the 3 variables are the most suitable ones, nor that the FCM-FN based models overwhelm the earlier models in terms of performances other than interpretability and accuracy (<i>SMAPE<\/i> and <i>R<\/i><sup>2<\/sup>). Our results showed that FN based models improves the interpretability and accuracy significantly. Notwithstanding its limitation, it is possible to improve the accuracy by implementing more linguistic terms and more accurate subjective appraisals in FCM-FN models. Despite its preliminary character, FCM-FN model is effective for prediction.<br \/>To optimize this model in the future, we would like to conduct experiment from the following four aspects: First, integrate type-2 fuzzy system and put it forward in FCM-FN models. Second, integrate other optimization methods, such as GA and \u201c<b>Patternsearch<\/b>\u201d. As an important optimization method in function \u201c<b>tunefis<\/b>\u201d, \u201c<b>Patternsearch<\/b>\u201d performs better for small parameter ranges since they are local optimizers. \u201c<b>Patternsearch<\/b>\u201d may produce faster convergence compared to particleswarm as the tradeoff for accuracy. Third, change the number of linguistic terms or the number of clusters in FCM to explore their influence on accuracy. Forth, integrate newly fuzzy clustering methods to explore their influence on accuracy and interpretability.<br \/>The datasets used and analyzed during the current study are available from three different ways. Three datasets (white wine, red wine and concrete) are available from the UCI Machine Learning Repository (<a href=\"http:\/\/archive.ics.uci.edu\/ml\/index.php\">http:\/\/archive.ics.uci.edu\/ml\/index.php<\/a>), the datasets (boston and diabetes) are available from kaggle website (<a href=\"https:\/\/www.kaggle.com\/datasets\/\">https:\/\/www.kaggle.com\/datasets\/<\/a>) and the QUES dataset originally published by Li and Henry<sup><a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 43\" title=\"Li, W. &amp; Henry, S. Object-oriented metrics that predict maintainability. J. Syst. Softw. 23, 111\u2013122. &#xA;                  https:\/\/doi.org\/10.1016\/0164-1212(93)90077-B&#xA;                  &#xA;                 (1993).\" href=\"\/articles\/s41598-022-20015-y#ref-CR43\" id=\"ref-link-section-d104546385e14080\">43<\/a><\/sup>.<br \/>The MATLAB source code for this project is available upon reasonable request. Please contact Xiaowei Wang (xiaowwei.wang@outlook.com).<br \/>Wang, G., Wang, H. &amp; Long, Z. Norm approximation of mamdani fuzzy system to a class of integrable functions. <i>Int. J. Fuzzy Syst.<\/i> <b>23<\/b>, 833\u2013848. <a href=\"https:\/\/doi.org\/10.1007\/s40815-020-01008-3\">https:\/\/doi.org\/10.1007\/s40815-020-01008-3<\/a> (2021).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1007\/s40815-020-01008-3\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1007%2Fs40815-020-01008-3\" aria-label=\"Article reference 1\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 1\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Norm%20approximation%20of%20mamdani%20fuzzy%20system%20to%20a%20class%20of%20integrable%20functions&amp;journal=Int.%20J.%20Fuzzy%20Syst.&amp;doi=10.1007%2Fs40815-020-01008-3&amp;volume=23&amp;pages=833-848&amp;publication_year=2021&amp;author=Wang%2CG&amp;author=Wang%2CH&amp;author=Long%2CZ\">                     Google Scholar<\/a>\u00a0                 <br \/>Tsekouras, G., Sarimveis, H., Kavakli, E. &amp; Bafas, G. A hierarchical fuzzy-clustering approach to fuzzy modeling. <i>Fuzzy Sets Syst.<\/i> <b>150<\/b>, 245\u2013266. <a href=\"https:\/\/doi.org\/10.1016\/j.fss.2004.04.013\">https:\/\/doi.org\/10.1016\/j.fss.2004.04.013<\/a> (2005).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"mathscinet reference\" href=\"http:\/\/www.ams.org\/mathscinet-getitem?mr=2134621\" aria-label=\"MathSciNet reference 2\">MathSciNet<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.fss.2004.04.013\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.fss.2004.04.013\" aria-label=\"Article reference 2\">Article<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"math reference\" href=\"http:\/\/www.emis.de\/MATH-item?1142.93311\" aria-label=\"MATH reference 2\">MATH<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 2\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=A%20hierarchical%20fuzzy-clustering%20approach%20to%20fuzzy%20modeling&amp;journal=Fuzzy%20Sets%20Syst.&amp;doi=10.1016%2Fj.fss.2004.04.013&amp;volume=150&amp;pages=245-266&amp;publication_year=2005&amp;author=Tsekouras%2CG&amp;author=Sarimveis%2CH&amp;author=Kavakli%2CE&amp;author=Bafas%2CG\">                     Google Scholar<\/a>\u00a0                 <br \/>Zhang, Y., Wu, X. B., Xing, Z. Y. &amp; Hu, W. L. On generating interpretable and precise fuzzy systems based on Pareto multi-objective cooperative co-evolutionary algorithm. <i>Appl. Soft Comput.<\/i> <b>11<\/b>, 1284\u20131294. <a href=\"https:\/\/doi.org\/10.1016\/j.asoc.2010.03.005\">https:\/\/doi.org\/10.1016\/j.asoc.2010.03.005<\/a> (2011).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.asoc.2010.03.005\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.asoc.2010.03.005\" aria-label=\"Article reference 3\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 3\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=On%20generating%20interpretable%20and%20precise%20fuzzy%20systems%20based%20on%20Pareto%20multi-objective%20cooperative%20co-evolutionary%20algorithm&amp;journal=Appl.%20Soft%20Comput.&amp;doi=10.1016%2Fj.asoc.2010.03.005&amp;volume=11&amp;pages=1284-1294&amp;publication_year=2011&amp;author=Zhang%2CY&amp;author=Wu%2CXB&amp;author=Xing%2CZY&amp;author=Hu%2CWL\">                     Google Scholar<\/a>\u00a0                 <br \/>Zador, A. M. A critique of pure learning and what artificial neural networks can learn from animal brains. <i>Nat. Commun.<\/i> <b>10<\/b>, 1\u20137. <a href=\"https:\/\/doi.org\/10.1038\/s41467-019-11786-6\">https:\/\/doi.org\/10.1038\/s41467-019-11786-6<\/a> (2019).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"ads reference\" href=\"http:\/\/adsabs.harvard.edu\/cgi-bin\/nph-data_query?link_type=ABSTRACT&amp;bibcode=2019OptCo.433....1Z\" aria-label=\"ADS reference 4\">ADS<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"cas reference\" href=\"\/articles\/cas-redirect\/1:CAS:528:DC%2BC1MXhs1entb7M\" aria-label=\"CAS reference 4\">CAS<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1038\/s41467-019-11786-6\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1038%2Fs41467-019-11786-6\" aria-label=\"Article reference 4\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 4\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=A%20critique%20of%20pure%20learning%20and%20what%20artificial%20neural%20networks%20can%20learn%20from%20animal%20brains&amp;journal=Nat.%20Commun.&amp;doi=10.1038%2Fs41467-019-11786-6&amp;volume=10&amp;pages=1-7&amp;publication_year=2019&amp;author=Zador%2CAM\">                     Google Scholar<\/a>\u00a0                 <br \/>Dick, S., Meeks, A., Last, M., Bunke, H. &amp; Kandel, A. Data mining in software metrics databases. <i>Fuzzy Sets Syst.<\/i> <b>145<\/b>, 81\u2013110. <a href=\"https:\/\/doi.org\/10.1016\/j.fss.2003.10.006\">https:\/\/doi.org\/10.1016\/j.fss.2003.10.006<\/a> (2004).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"mathscinet reference\" href=\"http:\/\/www.ams.org\/mathscinet-getitem?mr=2061942\" aria-label=\"MathSciNet reference 5\">MathSciNet<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.fss.2003.10.006\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.fss.2003.10.006\" aria-label=\"Article reference 5\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 5\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Data%20mining%20in%20software%20metrics%20databases&amp;journal=Fuzzy%20Sets%20Syst.&amp;doi=10.1016%2Fj.fss.2003.10.006&amp;volume=145&amp;pages=81-110&amp;publication_year=2004&amp;author=Dick%2CS&amp;author=Meeks%2CA&amp;author=Last%2CM&amp;author=Bunke%2CH&amp;author=Kandel%2CA\">                     Google Scholar<\/a>\u00a0                 <br \/>Guillaume, S. Designing fuzzy inference systems from data: An interpretability-oriented review. <i>Fuzzy Syst. IEEE Trans.<\/i> <b>9<\/b>, 426\u2013443. <a href=\"https:\/\/doi.org\/10.1109\/91.928739\">https:\/\/doi.org\/10.1109\/91.928739<\/a> (2001).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1109\/91.928739\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1109%2F91.928739\" aria-label=\"Article reference 6\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 6\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Designing%20fuzzy%20inference%20systems%20from%20data%3A%20An%20interpretability-oriented%20review&amp;journal=Fuzzy%20Syst.%20IEEE%20Trans.&amp;doi=10.1109%2F91.928739&amp;volume=9&amp;pages=426-443&amp;publication_year=2001&amp;author=Guillaume%2CS\">                     Google Scholar<\/a>\u00a0                 <br \/>Jiang, G. C. &amp; Sun, Z. J. Improvement on a fuzzy C-means algorithm based on genetic algorithm. <i>Appl. Mech. Mater.<\/i> <b>614<\/b>, 385\u2013388. <a href=\"https:\/\/doi.org\/10.4028\/www.scientific.net\/AMM.614.385\">https:\/\/doi.org\/10.4028\/www.scientific.net\/AMM.614.385<\/a> (2014).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.4028\/www.scientific.net\/AMM.614.385\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.4028%2Fwww.scientific.net%2FAMM.614.385\" aria-label=\"Article reference 7\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 7\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Improvement%20on%20a%20fuzzy%20C-means%20algorithm%20based%20on%20genetic%20algorithm&amp;journal=Appl.%20Mech.%20Mater.&amp;doi=10.4028%2Fwww.scientific.net%2FAMM.614.385&amp;volume=614&amp;pages=385-388&amp;publication_year=2014&amp;author=Jiang%2CGC&amp;author=Sun%2CZJ\">                     Google Scholar<\/a>\u00a0                 <br \/>Ding, Y. &amp; Fu, X. Kernel-based fuzzy c-means clustering algorithm based on genetic algorithm. <i>Neurocomputing<\/i> <b>188<\/b>, 233\u2013238. <a href=\"https:\/\/doi.org\/10.1016\/j.neucom.2015.01.106\">https:\/\/doi.org\/10.1016\/j.neucom.2015.01.106<\/a> (2016).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.neucom.2015.01.106\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.neucom.2015.01.106\" aria-label=\"Article reference 8\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 8\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Kernel-based%20fuzzy%20c-means%20clustering%20algorithm%20based%20on%20genetic%20algorithm&amp;journal=Neurocomputing&amp;doi=10.1016%2Fj.neucom.2015.01.106&amp;volume=188&amp;pages=233-238&amp;publication_year=2016&amp;author=Ding%2CY&amp;author=Fu%2CX\">                     Google Scholar<\/a>\u00a0                 <br \/>Silva-Filho, T. M., Pimentel, B. A., Souza, R. M. C. R. &amp; Oliveira, A. L. I. Hybrid methods for fuzzy clustering based on fuzzy c-means and improved particle swarm optimization. <i>Expert Syst. Appl.<\/i> <b>42<\/b>, 6315\u20136328. <a href=\"https:\/\/doi.org\/10.1016\/j.eswa.2015.04.032\">https:\/\/doi.org\/10.1016\/j.eswa.2015.04.032<\/a> (2015).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.eswa.2015.04.032\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.eswa.2015.04.032\" aria-label=\"Article reference 9\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 9\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Hybrid%20methods%20for%20fuzzy%20clustering%20based%20on%20fuzzy%20c-means%20and%20improved%20particle%20swarm%20optimization&amp;journal=Expert%20Syst.%20Appl.&amp;doi=10.1016%2Fj.eswa.2015.04.032&amp;volume=42&amp;pages=6315-6328&amp;publication_year=2015&amp;author=Silva-Filho%2CTM&amp;author=Pimentel%2CBA&amp;author=Souza%2CRMCR&amp;author=Oliveira%2CALI\">                     Google Scholar<\/a>\u00a0                 <br \/>Kuo, R. J., Lin, T. C., Zulvia, F. E. &amp; Tsai, C. Y. A hybrid metaheuristic and kernel intuitionistic fuzzy c-means algorithm for cluster analysis. <i>Appl. Soft Comput.<\/i> <b>67<\/b>, 299\u2013308. <a href=\"https:\/\/doi.org\/10.1016\/j.asoc.2018.02.039\">https:\/\/doi.org\/10.1016\/j.asoc.2018.02.039<\/a> (2018).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.asoc.2018.02.039\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.asoc.2018.02.039\" aria-label=\"Article reference 10\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 10\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=A%20hybrid%20metaheuristic%20and%20kernel%20intuitionistic%20fuzzy%20c-means%20algorithm%20for%20cluster%20analysis&amp;journal=Appl.%20Soft%20Comput.&amp;doi=10.1016%2Fj.asoc.2018.02.039&amp;volume=67&amp;pages=299-308&amp;publication_year=2018&amp;author=Kuo%2CRJ&amp;author=Lin%2CTC&amp;author=Zulvia%2CFE&amp;author=Tsai%2CCY\">                     Google Scholar<\/a>\u00a0                 <br \/>Izakian, H. &amp; Abraham, A. Fuzzy C-means and fuzzy swarm for fuzzy clustering problem. <i>Expert Syst. Appl.<\/i> <b>38<\/b>, 1835\u20131838. <a href=\"https:\/\/doi.org\/10.1016\/j.eswa.2010.07.112\">https:\/\/doi.org\/10.1016\/j.eswa.2010.07.112<\/a> (2011).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.eswa.2010.07.112\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.eswa.2010.07.112\" aria-label=\"Article reference 11\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 11\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Fuzzy%20C-means%20and%20fuzzy%20swarm%20for%20fuzzy%20clustering%20problem&amp;journal=Expert%20Syst.%20Appl.&amp;doi=10.1016%2Fj.eswa.2010.07.112&amp;volume=38&amp;pages=1835-1838&amp;publication_year=2011&amp;author=Izakian%2CH&amp;author=Abraham%2CA\">                     Google Scholar<\/a>\u00a0                 <br \/>Hu, J. &amp; Yin, H. FCM clustering algorithm based on PSO-TVAC algorithm with adaptively weighted centers. <i>Adv. Appl. Math.<\/i> <b>10<\/b>, 953\u2013962. <a href=\"https:\/\/doi.org\/10.12677\/AAM.2021.104104\">https:\/\/doi.org\/10.12677\/AAM.2021.104104<\/a> (2021).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.12677\/AAM.2021.104104\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.12677%2FAAM.2021.104104\" aria-label=\"Article reference 12\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 12\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=FCM%20clustering%20algorithm%20based%20on%20PSO-TVAC%20algorithm%20with%20adaptively%20weighted%20centers&amp;journal=Adv.%20Appl.%20Math.&amp;doi=10.12677%2FAAM.2021.104104&amp;volume=10&amp;pages=953-962&amp;publication_year=2021&amp;author=Hu%2CJ&amp;author=Yin%2CH\">                     Google Scholar<\/a>\u00a0                 <br \/>Zhang, Y. C., Xiong, X. &amp; Zhang, Q. D. An improved self-adaptive PSO algorithm with detection function for multimodal function optimization problems. <i>Math. Probl. Eng.<\/i> <b>657\u2013675<\/b>, 2013. <a href=\"https:\/\/doi.org\/10.1155\/2013\/716952\">https:\/\/doi.org\/10.1155\/2013\/716952<\/a> (2013).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"mathscinet reference\" href=\"http:\/\/www.ams.org\/mathscinet-getitem?mr=3148420\" aria-label=\"MathSciNet reference 13\">MathSciNet<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1155\/2013\/716952\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1155%2F2013%2F716952\" aria-label=\"Article reference 13\">Article<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"math reference\" href=\"http:\/\/www.emis.de\/MATH-item?1299.90430\" aria-label=\"MATH reference 13\">MATH<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 13\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=An%20improved%20self-adaptive%20PSO%20algorithm%20with%20detection%20function%20for%20multimodal%20function%20optimization%20problems&amp;journal=Math.%20Probl.%20Eng.&amp;doi=10.1155%2F2013%2F716952&amp;volume=657%E2%80%93675&amp;publication_year=2013&amp;author=Zhang%2CYC&amp;author=Xiong%2CX&amp;author=Zhang%2CQD\">                     Google Scholar<\/a>\u00a0                 <br \/>Alam, S., Dobbie, G., Koh, Y. S., Riddle, P. &amp; Rehman, S. U. Research on particle swarm optimization based clustering: A systematic review of literature and techniques. <i>Swarm Evol. Comput.<\/i> <b>17<\/b>, 1\u201313. <a href=\"https:\/\/doi.org\/10.1016\/j.swevo.2014.02.001\">https:\/\/doi.org\/10.1016\/j.swevo.2014.02.001<\/a> (2014).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.swevo.2014.02.001\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.swevo.2014.02.001\" aria-label=\"Article reference 14\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 14\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Research%20on%20particle%20swarm%20optimization%20based%20clustering%3A%20A%20systematic%20review%20of%20literature%20and%20techniques&amp;journal=Swarm%20Evol.%20Comput.&amp;doi=10.1016%2Fj.swevo.2014.02.001&amp;volume=17&amp;pages=1-13&amp;publication_year=2014&amp;author=Alam%2CS&amp;author=Dobbie%2CG&amp;author=Koh%2CYS&amp;author=Riddle%2CP&amp;author=Rehman%2CSU\">                     Google Scholar<\/a>\u00a0                 <br \/>Chen, S., Xu, Z. &amp; Tang, Y. A hybrid clustering algorithm based on fuzzy c-means and improved particle swarm optimization. <i>Arab. J. Sci. Eng.<\/i> <b>39<\/b>, 8875\u20138887. <a href=\"https:\/\/doi.org\/10.1007\/s13369-014-1424-9\">https:\/\/doi.org\/10.1007\/s13369-014-1424-9<\/a> (2014).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"mathscinet reference\" href=\"http:\/\/www.ams.org\/mathscinet-getitem?mr=3282823\" aria-label=\"MathSciNet reference 15\">MathSciNet<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1007\/s13369-014-1424-9\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1007%2Fs13369-014-1424-9\" aria-label=\"Article reference 15\">Article<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"math reference\" href=\"http:\/\/www.emis.de\/MATH-item?06882957\" aria-label=\"MATH reference 15\">MATH<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 15\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=A%20hybrid%20clustering%20algorithm%20based%20on%20fuzzy%20c-means%20and%20improved%20particle%20swarm%20optimization&amp;journal=Arab.%20J.%20Sci.%20Eng.&amp;doi=10.1007%2Fs13369-014-1424-9&amp;volume=39&amp;pages=8875-8887&amp;publication_year=2014&amp;author=Chen%2CS&amp;author=Xu%2CZ&amp;author=Tang%2CY\">                     Google Scholar<\/a>\u00a0                 <br \/>Ahmed, M. A. &amp; Al-Jamimi, H. A. Machine learning approaches for predicting software maintainability: A fuzzy-based transparent model. <i>IET Softw.<\/i> <b>7<\/b>, 317\u2013326. <a href=\"https:\/\/doi.org\/10.1049\/iet-sen.2013.0046\">https:\/\/doi.org\/10.1049\/iet-sen.2013.0046<\/a> (2013).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1049\/iet-sen.2013.0046\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1049%2Fiet-sen.2013.0046\" aria-label=\"Article reference 16\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 16\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Machine%20learning%20approaches%20for%20predicting%20software%20maintainability%3A%20A%20fuzzy-based%20transparent%20model&amp;journal=IET%20Softw.&amp;doi=10.1049%2Fiet-sen.2013.0046&amp;volume=7&amp;pages=317-326&amp;publication_year=2013&amp;author=Ahmed%2CMA&amp;author=Al-Jamimi%2CHA\">                     Google Scholar<\/a>\u00a0                 <br \/>Wang, L.-X. &amp; Mendel, J. M. Generating fuzzy rules by learning from examples. <i>IEEE Trans. Syst. Man Cybern.<\/i> <b>22<\/b>, 1414\u20131427. <a href=\"https:\/\/doi.org\/10.1109\/21.199466\">https:\/\/doi.org\/10.1109\/21.199466<\/a> (1992).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"mathscinet reference\" href=\"http:\/\/www.ams.org\/mathscinet-getitem?mr=1212464\" aria-label=\"MathSciNet reference 17\">MathSciNet<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1109\/21.199466\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1109%2F21.199466\" aria-label=\"Article reference 17\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 17\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Generating%20fuzzy%20rules%20by%20learning%20from%20examples&amp;journal=IEEE%20Trans.%20Syst.%20Man%20Cybern.&amp;doi=10.1109%2F21.199466&amp;volume=22&amp;pages=1414-1427&amp;publication_year=1992&amp;author=Wang%2CL-X&amp;author=Mendel%2CJM\">                     Google Scholar<\/a>\u00a0                 <br \/>Wang, X., Gegov, A., Farzad, A., Chen, Y. &amp; Hu, Q. Fuzzy network based framework for software maintainability prediction. <i>Internat. J. Uncertain. Fuzziness Knowl.-Based Syst.<\/i> <b>27<\/b>, 841\u2013862. <a href=\"https:\/\/doi.org\/10.1142\/s0218488519500375\">https:\/\/doi.org\/10.1142\/s0218488519500375<\/a> (2019).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"cas reference\" href=\"\/articles\/cas-redirect\/1:CAS:528:DC%2BC1MXht1Ojsr7P\" aria-label=\"CAS reference 18\">CAS<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1142\/s0218488519500375\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1142%2Fs0218488519500375\" aria-label=\"Article reference 18\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 18\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Fuzzy%20network%20based%20framework%20for%20software%20maintainability%20prediction&amp;journal=Internat.%20J.%20Uncertain.%20Fuzziness%20Knowl.-Based%20Syst.&amp;doi=10.1142%2Fs0218488519500375&amp;volume=27&amp;pages=841-862&amp;publication_year=2019&amp;author=Wang%2CX&amp;author=Gegov%2CA&amp;author=Farzad%2CA&amp;author=Chen%2CY&amp;author=Hu%2CQ\">                     Google Scholar<\/a>\u00a0                 <br \/>Bucolo, M., Fortuna, L. &amp; Rosa, M. L. Complex dynamics through fuzzy chains. <i>IEEE Trans. Fuzzy Syst.<\/i> <b>12<\/b>, 289\u2013295. <a href=\"https:\/\/doi.org\/10.1109\/TFUZZ.2004.825969\">https:\/\/doi.org\/10.1109\/TFUZZ.2004.825969<\/a> (2004).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1109\/TFUZZ.2004.825969\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1109%2FTFUZZ.2004.825969\" aria-label=\"Article reference 19\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 19\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Complex%20dynamics%20through%20fuzzy%20chains&amp;journal=IEEE%20Trans.%20Fuzzy%20Syst.&amp;doi=10.1109%2FTFUZZ.2004.825969&amp;volume=12&amp;pages=289-295&amp;publication_year=2004&amp;author=Bucolo%2CM&amp;author=Fortuna%2CL&amp;author=Rosa%2CML\">                     Google Scholar<\/a>\u00a0                 <br \/>Lee, M.-L., Chung, H.-Y. &amp; Yu, F.-M. Modeling of hierarchical fuzzy systems. <i>Fuzzy Sets Syst.<\/i> <b>138<\/b>, 343\u2013361. <a href=\"https:\/\/doi.org\/10.1016\/S0165-0114(02)00517-1\">https:\/\/doi.org\/10.1016\/S0165-0114(02)00517-1<\/a> (2003).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"mathscinet reference\" href=\"http:\/\/www.ams.org\/mathscinet-getitem?mr=1999626\" aria-label=\"MathSciNet reference 20\">MathSciNet<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/S0165-0114(02)00517-1\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2FS0165-0114%2802%2900517-1\" aria-label=\"Article reference 20\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 20\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Modeling%20of%20hierarchical%20fuzzy%20systems&amp;journal=Fuzzy%20Sets%20Syst.&amp;doi=10.1016%2FS0165-0114%2802%2900517-1&amp;volume=138&amp;pages=343-361&amp;publication_year=2003&amp;author=Lee%2CM-L&amp;author=Chung%2CH-Y&amp;author=Yu%2CF-M\">                     Google Scholar<\/a>\u00a0                 <br \/>Joo, M. G. &amp; Lee, J. S. A class of hierarchical fuzzy systems with constraints on the fuzzy rules. <i>IEEE Trans. Fuzzy Syst.<\/i> <b>13<\/b>, 194\u2013203. <a href=\"https:\/\/doi.org\/10.1109\/TFUZZ.2004.840096\">https:\/\/doi.org\/10.1109\/TFUZZ.2004.840096<\/a> (2005).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1109\/TFUZZ.2004.840096\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1109%2FTFUZZ.2004.840096\" aria-label=\"Article reference 21\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 21\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=A%20class%20of%20hierarchical%20fuzzy%20systems%20with%20constraints%20on%20the%20fuzzy%20rules&amp;journal=IEEE%20Trans.%20Fuzzy%20Syst.&amp;doi=10.1109%2FTFUZZ.2004.840096&amp;volume=13&amp;pages=194-203&amp;publication_year=2005&amp;author=Joo%2CMG&amp;author=Lee%2CJS\">                     Google Scholar<\/a>\u00a0                 <br \/>Gegov, A., Arabikhan, F. &amp; Petrov, N. Linguistic composition based modelling by fuzzy networks with modular rule bases. <i>Fuzzy Sets Syst.<\/i> <b>269<\/b>, 1\u201329. <a href=\"https:\/\/doi.org\/10.1016\/j.fss.2014.06.014\">https:\/\/doi.org\/10.1016\/j.fss.2014.06.014<\/a> (2014).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"mathscinet reference\" href=\"http:\/\/www.ams.org\/mathscinet-getitem?mr=3324001\" aria-label=\"MathSciNet reference 22\">MathSciNet<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.fss.2014.06.014\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.fss.2014.06.014\" aria-label=\"Article reference 22\">Article<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"math reference\" href=\"http:\/\/www.emis.de\/MATH-item?1361.93006\" aria-label=\"MATH reference 22\">MATH<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 22\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Linguistic%20composition%20based%20modelling%20by%20fuzzy%20networks%20with%20modular%20rule%20bases&amp;journal=Fuzzy%20Sets%20Syst.&amp;doi=10.1016%2Fj.fss.2014.06.014&amp;volume=269&amp;pages=1-29&amp;publication_year=2014&amp;author=Gegov%2CA&amp;author=Arabikhan%2CF&amp;author=Petrov%2CN\">                     Google Scholar<\/a>\u00a0                 <br \/>Yaakob, A. M., Gegov, A. &amp; Rahman, S. Selection of alternatives using fuzzy networks with rule base aggregation. <i>Fuzzy Sets Syst.<\/i> <b>341<\/b>, 123\u2013144. <a href=\"https:\/\/doi.org\/10.1016\/j.fss.2017.05.027\">https:\/\/doi.org\/10.1016\/j.fss.2017.05.027<\/a> (2018).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"mathscinet reference\" href=\"http:\/\/www.ams.org\/mathscinet-getitem?mr=3787607\" aria-label=\"MathSciNet reference 23\">MathSciNet<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.fss.2017.05.027\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.fss.2017.05.027\" aria-label=\"Article reference 23\">Article<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"math reference\" href=\"http:\/\/www.emis.de\/MATH-item?1397.91155\" aria-label=\"MATH reference 23\">MATH<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 23\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Selection%20of%20alternatives%20using%20fuzzy%20networks%20with%20rule%20base%20aggregation&amp;journal=Fuzzy%20Sets%20Syst.&amp;doi=10.1016%2Fj.fss.2017.05.027&amp;volume=341&amp;pages=123-144&amp;publication_year=2018&amp;author=Yaakob%2CAM&amp;author=Gegov%2CA&amp;author=Rahman%2CS\">                     Google Scholar<\/a>\u00a0                 <br \/>Gegov, A., Sanders, D. &amp; Vatchova, B. Mamdani fuzzy networks with feedforward rule bases for complex systems modelling. <i>J. Intell. Fuzzy Syst.<\/i> <b>30<\/b>, 2623\u20132637. <a href=\"https:\/\/doi.org\/10.3233\/IFS-151911\">https:\/\/doi.org\/10.3233\/IFS-151911<\/a> (2016).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.3233\/IFS-151911\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.3233%2FIFS-151911\" aria-label=\"Article reference 24\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 24\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Mamdani%20fuzzy%20networks%20with%20feedforward%20rule%20bases%20for%20complex%20systems%20modelling&amp;journal=J.%20Intell.%20Fuzzy%20Syst.&amp;doi=10.3233%2FIFS-151911&amp;volume=30&amp;pages=2623-2637&amp;publication_year=2016&amp;author=Gegov%2CA&amp;author=Sanders%2CD&amp;author=Vatchova%2CB\">                     Google Scholar<\/a>\u00a0                 <br \/>Gegov, A. <i>Fuzzy Networks for Complex Systems: A Modular Rule Base Approach<\/i>. <a href=\"https:\/\/doi.org\/10.1007\/978-3-642-15600-7\">https:\/\/doi.org\/10.1007\/978-3-642-15600-7<\/a> (Springer, 2010).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1007\/978-3-642-15600-7\" data-track-action=\"book reference\" href=\"https:\/\/doi.org\/10.1007%2F978-3-642-15600-7\" aria-label=\"Book reference 25\">Book<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"math reference\" href=\"http:\/\/www.emis.de\/MATH-item?1202.68027\" aria-label=\"MATH reference 25\">MATH<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 25\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Fuzzy%20Networks%20for%20Complex%20Systems%3A%20A%20Modular%20Rule%20Base%20Approach&amp;publication_year=2010&amp;author=Gegov%2CA\">                     Google Scholar<\/a>\u00a0                 <br \/>Mendel, J. M. <i>Uncertain Rule-Based Fuzzy Logic Systems: Introduction and New Directions<\/i> 2nd edn. (Springer, 2017).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1007\/978-3-319-51370-6\" data-track-action=\"book reference\" href=\"https:\/\/doi.org\/10.1007%2F978-3-319-51370-6\" aria-label=\"Book reference 26\">Book<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 26\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Uncertain%20Rule-Based%20Fuzzy%20Logic%20Systems%3A%20Introduction%20and%20New%20Directions&amp;publication_year=2017&amp;author=Mendel%2CJM\">                     Google Scholar<\/a>\u00a0                 <br \/>Zadeh, L. A. Outline of a new approach to the analysis of complex systems and decision processes. <i>IEEE Trans. Syst. Man Cybern.<\/i> <b>3<\/b>, 28\u201344. <a href=\"https:\/\/doi.org\/10.1109\/TSMC.1973.5408575\">https:\/\/doi.org\/10.1109\/TSMC.1973.5408575<\/a> (1973).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"mathscinet reference\" href=\"http:\/\/www.ams.org\/mathscinet-getitem?mr=309582\" aria-label=\"MathSciNet reference 27\">MathSciNet<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1109\/TSMC.1973.5408575\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1109%2FTSMC.1973.5408575\" aria-label=\"Article reference 27\">Article<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"math reference\" href=\"http:\/\/www.emis.de\/MATH-item?0273.93002\" aria-label=\"MATH reference 27\">MATH<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 27\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Outline%20of%20a%20new%20approach%20to%20the%20analysis%20of%20complex%20systems%20and%20decision%20processes&amp;journal=IEEE%20Trans.%20Syst.%20Man%20Cybern.&amp;doi=10.1109%2FTSMC.1973.5408575&amp;volume=3&amp;pages=28-44&amp;publication_year=1973&amp;author=Zadeh%2CLA\">                     Google Scholar<\/a>\u00a0                 <br \/>Zadeh, L. A. Fuzzy sets. <i>Inf. Control<\/i> <b>8<\/b>, 338\u2013353. <a href=\"https:\/\/doi.org\/10.1016\/S0019-9958(65)90241-X\">https:\/\/doi.org\/10.1016\/S0019-9958(65)90241-X<\/a> (1965).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/S0019-9958(65)90241-X\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2FS0019-9958%2865%2990241-X\" aria-label=\"Article reference 28\">Article<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"math reference\" href=\"http:\/\/www.emis.de\/MATH-item?0139.24606\" aria-label=\"MATH reference 28\">MATH<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 28\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Fuzzy%20sets&amp;journal=Inf.%20Control&amp;doi=10.1016%2FS0019-9958%2865%2990241-X&amp;volume=8&amp;pages=338-353&amp;publication_year=1965&amp;author=Zadeh%2CLA\">                     Google Scholar<\/a>\u00a0                 <br \/>Mamdani, E. H. Application of fuzzy logic to approximate reasoning using linguistic synthesis. <i>Comput. IEEE Trans.<\/i> <b>100<\/b>, 1182\u20131191. <a href=\"https:\/\/doi.org\/10.1109\/TC.1977.1674779\">https:\/\/doi.org\/10.1109\/TC.1977.1674779<\/a> (1977).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1109\/TC.1977.1674779\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1109%2FTC.1977.1674779\" aria-label=\"Article reference 29\">Article<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"math reference\" href=\"http:\/\/www.emis.de\/MATH-item?0397.94025\" aria-label=\"MATH reference 29\">MATH<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 29\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Application%20of%20fuzzy%20logic%20to%20approximate%20reasoning%20using%20linguistic%20synthesis&amp;journal=Comput.%20IEEE%20Trans.&amp;doi=10.1109%2FTC.1977.1674779&amp;volume=100&amp;pages=1182-1191&amp;publication_year=1977&amp;author=Mamdani%2CEH\">                     Google Scholar<\/a>\u00a0                 <br \/>Full, J. C. B. R. E. W. FCM: The fuzzy c -means clustering algorithm. <i>Comput. Geosci.<\/i> <b>10<\/b>, 191\u2013203. <a href=\"https:\/\/doi.org\/10.1016\/0098-3004(84)90020-7\">https:\/\/doi.org\/10.1016\/0098-3004(84)90020-7<\/a> (1984).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"ads reference\" href=\"http:\/\/adsabs.harvard.edu\/cgi-bin\/nph-data_query?link_type=ABSTRACT&amp;bibcode=1984CG.....10..191B\" aria-label=\"ADS reference 30\">ADS<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/0098-3004(84)90020-7\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2F0098-3004%2884%2990020-7\" aria-label=\"Article reference 30\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 30\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=FCM%3A%20The%20fuzzy%20c%20-means%20clustering%20algorithm&amp;journal=Comput.%20Geosci.&amp;doi=10.1016%2F0098-3004%2884%2990020-7&amp;volume=10&amp;pages=191-203&amp;publication_year=1984&amp;author=Full%2CJCBREW\">                     Google Scholar<\/a>\u00a0                 <br \/>Askari, S., Montazerin, N. &amp; Zarandi, M. F. Generalized possibilistic fuzzy c-means with novel cluster validity indices for clustering noisy data. <i>Appl. Soft Comput.<\/i> <b>53<\/b>, 262\u2013283. <a href=\"https:\/\/doi.org\/10.1016\/j.asoc.2016.12.049\">https:\/\/doi.org\/10.1016\/j.asoc.2016.12.049<\/a> (2017).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.asoc.2016.12.049\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.asoc.2016.12.049\" aria-label=\"Article reference 31\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 31\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Generalized%20possibilistic%20fuzzy%20c-means%20with%20novel%20cluster%20validity%20indices%20for%20clustering%20noisy%20data&amp;journal=Appl.%20Soft%20Comput.&amp;doi=10.1016%2Fj.asoc.2016.12.049&amp;volume=53&amp;pages=262-283&amp;publication_year=2017&amp;author=Askari%2CS&amp;author=Montazerin%2CN&amp;author=Zarandi%2CMF\">                     Google Scholar<\/a>\u00a0                 <br \/>Rubio, E. <i>et al.<\/i> An extension of the fuzzy possibilistic clustering algorithm using type-2 fuzzy logic techniques. <i>Adv. Fuzzy Syst.<\/i> <b>2017<\/b>, 53. <a href=\"https:\/\/doi.org\/10.1155\/2017\/7094046\">https:\/\/doi.org\/10.1155\/2017\/7094046<\/a> (2017).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1155\/2017\/7094046\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1155%2F2017%2F7094046\" aria-label=\"Article reference 32\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 32\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=An%20extension%20of%20the%20fuzzy%20possibilistic%20clustering%20algorithm%20using%20type-2%20fuzzy%20logic%20techniques&amp;journal=Adv.%20Fuzzy%20Syst.&amp;doi=10.1155%2F2017%2F7094046&amp;volume=2017&amp;publication_year=2017&amp;author=Rubio%2CE\">                     Google Scholar<\/a>\u00a0                 <br \/>Mansoureh, N., Mohammad HosseinFazel, Z. &amp; Susan, B. A fuzzy cluster-validity index based on the topology structure and node attribute in complex networks. <i>Expert Syst. Appl.<\/i> <b>187<\/b>, 115913. <a href=\"https:\/\/doi.org\/10.1016\/j.eswa.2021.115913\">https:\/\/doi.org\/10.1016\/j.eswa.2021.115913<\/a> (2022).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.eswa.2021.115913\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.eswa.2021.115913\" aria-label=\"Article reference 33\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 33\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=A%20fuzzy%20cluster-validity%20index%20based%20on%20the%20topology%20structure%20and%20node%20attribute%20in%20complex%20networks&amp;journal=Expert%20Syst.%20Appl.&amp;doi=10.1016%2Fj.eswa.2021.115913&amp;volume=187&amp;publication_year=2022&amp;author=Mansoureh%2CN&amp;author=Mohammad%20HosseinFazel%2CZ&amp;author=Susan%2CB\">                     Google Scholar<\/a>\u00a0                 <br \/>Kaixin, Z., Yaping, D., Zhiyang, J. &amp; Ye, J. General fuzzy C-means clustering algorithm using Minkowski metric. <i>Signal Process.<\/i> <b>188<\/b>, 108161. <a href=\"https:\/\/doi.org\/10.1016\/j.sigpro.2021.108161\">https:\/\/doi.org\/10.1016\/j.sigpro.2021.108161<\/a> (2021).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.sigpro.2021.108161\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.sigpro.2021.108161\" aria-label=\"Article reference 34\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 34\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=General%20fuzzy%20C-means%20clustering%20algorithm%20using%20Minkowski%20metric&amp;journal=Signal%20Process.&amp;doi=10.1016%2Fj.sigpro.2021.108161&amp;volume=188&amp;publication_year=2021&amp;author=Kaixin%2CZ&amp;author=Yaping%2CD&amp;author=Zhiyang%2CJ&amp;author=Ye%2CJ\">                     Google Scholar<\/a>\u00a0                 <br \/>Filev, D. Fuzzy modeling of complex systems. <i>Int. J. Approx. Reason.<\/i> <b>5<\/b>, 281\u2013290. <a href=\"https:\/\/doi.org\/10.1016\/B978-1-4832-1450-4.50046-8\">https:\/\/doi.org\/10.1016\/B978-1-4832-1450-4.50046-8<\/a> (1991).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/B978-1-4832-1450-4.50046-8\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2FB978-1-4832-1450-4.50046-8\" aria-label=\"Article reference 35\">Article<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"math reference\" href=\"http:\/\/www.emis.de\/MATH-item?0738.93048\" aria-label=\"MATH reference 35\">MATH<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 35\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Fuzzy%20modeling%20of%20complex%20systems&amp;journal=Int.%20J.%20Approx.%20Reason.&amp;doi=10.1016%2FB978-1-4832-1450-4.50046-8&amp;volume=5&amp;pages=281-290&amp;publication_year=1991&amp;author=Filev%2CD\">                     Google Scholar<\/a>\u00a0                 <br \/>Fern\u00e1ndez Caballero, A. Contribution of fuzziness and uncertainty to modern artificial intelligence. <i>Fuzzy Sets Syst.<\/i> <b>160<\/b>, 129. <a href=\"https:\/\/doi.org\/10.1016\/j.fss.2008.05.002\">https:\/\/doi.org\/10.1016\/j.fss.2008.05.002<\/a> (2009).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"mathscinet reference\" href=\"http:\/\/www.ams.org\/mathscinet-getitem?mr=2473097\" aria-label=\"MathSciNet reference 36\">MathSciNet<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.fss.2008.05.002\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.fss.2008.05.002\" aria-label=\"Article reference 36\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 36\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Contribution%20of%20fuzziness%20and%20uncertainty%20to%20modern%20artificial%20intelligence&amp;journal=Fuzzy%20Sets%20Syst.&amp;doi=10.1016%2Fj.fss.2008.05.002&amp;volume=160&amp;publication_year=2009&amp;author=Fern%C3%A1ndez%20Caballero%2CA\">                     Google Scholar<\/a>\u00a0                 <br \/>Aja-Fernandez, S. &amp; Alberola-Lopez, C. Matrix modeling of hierarchical fuzzy systems. <i>IEEE Trans. Fuzzy Syst.<\/i> <b>16<\/b>, 585\u2013599. <a href=\"https:\/\/doi.org\/10.1109\/TFUZZ.2007.905905\">https:\/\/doi.org\/10.1109\/TFUZZ.2007.905905<\/a> (2008).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1109\/TFUZZ.2007.905905\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1109%2FTFUZZ.2007.905905\" aria-label=\"Article reference 37\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 37\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Matrix%20modeling%20of%20hierarchical%20fuzzy%20systems&amp;journal=IEEE%20Trans.%20Fuzzy%20Syst.&amp;doi=10.1109%2FTFUZZ.2007.905905&amp;volume=16&amp;pages=585-599&amp;publication_year=2008&amp;author=Aja-Fernandez%2CS&amp;author=Alberola-Lopez%2CC\">                     Google Scholar<\/a>\u00a0                 <br \/>Hall, L. O. Rule chaining in fuzzy expert systems. <i>IEEE Trans. Fuzzy Syst.<\/i> <b>9<\/b>, 822\u2013828. <a href=\"https:\/\/doi.org\/10.1109\/91.971731\">https:\/\/doi.org\/10.1109\/91.971731<\/a> (2001).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1109\/91.971731\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1109%2F91.971731\" aria-label=\"Article reference 38\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 38\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Rule%20chaining%20in%20fuzzy%20expert%20systems&amp;journal=IEEE%20Trans.%20Fuzzy%20Syst.&amp;doi=10.1109%2F91.971731&amp;volume=9&amp;pages=822-828&amp;publication_year=2001&amp;author=Hall%2CLO\">                     Google Scholar<\/a>\u00a0                 <br \/>Gegov, A., Petrov, N. &amp; Vatchova, B. In <i>Intelligent Systems (IS), 2010 5th IEEE International Conference<\/i> 197\u2013202 (IEEE, 2010).<br \/>Olugu, E. U. &amp; Wong, K. Y. An expert fuzzy rule-based system for closed-loop supply chain performance assessment in the automotive industry. <i>Expert Syst. Appl.<\/i> <b>39<\/b>, 375\u2013384. <a href=\"https:\/\/doi.org\/10.1016\/j.eswa.2011.07.026\">https:\/\/doi.org\/10.1016\/j.eswa.2011.07.026<\/a> (2012).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.eswa.2011.07.026\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.eswa.2011.07.026\" aria-label=\"Article reference 40\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 40\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=An%20expert%20fuzzy%20rule-based%20system%20for%20closed-loop%20supply%20chain%20performance%20assessment%20in%20the%20automotive%20industry&amp;journal=Expert%20Syst.%20Appl.&amp;doi=10.1016%2Fj.eswa.2011.07.026&amp;volume=39&amp;pages=375-384&amp;publication_year=2012&amp;author=Olugu%2CEU&amp;author=Wong%2CKY\">                     Google Scholar<\/a>\u00a0                 <br \/>Cortez, P., Cerdeira, A., Almeida, F., Matos, T. &amp; Reis, J. Modeling wine preferences by data mining from physicochemical properties. <i>Decis. Support Syst.<\/i> <b>47<\/b>, 547\u2013553. <a href=\"https:\/\/doi.org\/10.1016\/j.dss.2009.05.016\">https:\/\/doi.org\/10.1016\/j.dss.2009.05.016<\/a> (2009).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.dss.2009.05.016\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.dss.2009.05.016\" aria-label=\"Article reference 41\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 41\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Modeling%20wine%20preferences%20by%20data%20mining%20from%20physicochemical%20properties&amp;journal=Decis.%20Support%20Syst.&amp;doi=10.1016%2Fj.dss.2009.05.016&amp;volume=47&amp;pages=547-553&amp;publication_year=2009&amp;author=Cortez%2CP&amp;author=Cerdeira%2CA&amp;author=Almeida%2CF&amp;author=Matos%2CT&amp;author=Reis%2CJ\">                     Google Scholar<\/a>\u00a0                 <br \/>Yeh, I.-C. Modeling of strength of high-performance concrete using artificial neural networks. <i>Cem. Concr. Res.<\/i> <b>28<\/b>, 1797\u20131808. <a href=\"https:\/\/doi.org\/10.1016\/s0008-8846(98)00165-3\">https:\/\/doi.org\/10.1016\/s0008-8846(98)00165-3<\/a> (1998).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"link\" data-track-action=\"cas reference\" href=\"\/articles\/cas-redirect\/1:CAS:528:DyaK1MXivFOrug%3D%3D\" aria-label=\"CAS reference 42\">CAS<\/a>\u00a0     <a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/s0008-8846(98)00165-3\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fs0008-8846%2898%2900165-3\" aria-label=\"Article reference 42\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 42\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Modeling%20of%20strength%20of%20high-performance%20concrete%20using%20artificial%20neural%20networks&amp;journal=Cem.%20Concr.%20Res.&amp;doi=10.1016%2Fs0008-8846%2898%2900165-3&amp;volume=28&amp;pages=1797-1808&amp;publication_year=1998&amp;author=Yeh%2CI-C\">                     Google Scholar<\/a>\u00a0                 <br \/>Li, W. &amp; Henry, S. Object-oriented metrics that predict maintainability. <i>J. Syst. Softw.<\/i> <b>23<\/b>, 111\u2013122. <a href=\"https:\/\/doi.org\/10.1016\/0164-1212(93)90077-B\">https:\/\/doi.org\/10.1016\/0164-1212(93)90077-B<\/a> (1993).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/0164-1212(93)90077-B\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2F0164-1212%2893%2990077-B\" aria-label=\"Article reference 43\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 43\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Object-oriented%20metrics%20that%20predict%20maintainability&amp;journal=J.%20Syst.%20Softw.&amp;doi=10.1016%2F0164-1212%2893%2990077-B&amp;volume=23&amp;pages=111-122&amp;publication_year=1993&amp;author=Li%2CW&amp;author=Henry%2CS\">                     Google Scholar<\/a>\u00a0                 <br \/>Van den Broeke, M., De Baets, S., Vereecke, A., Baecke, P. &amp; Vanderheyden, K. Judgmental forecast adjustments over different time horizons. <i>Omega<\/i> <b>87<\/b>, 34\u201345. <a href=\"https:\/\/doi.org\/10.1016\/j.omega.2018.09.008\">https:\/\/doi.org\/10.1016\/j.omega.2018.09.008<\/a> (2019).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.omega.2018.09.008\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.omega.2018.09.008\" aria-label=\"Article reference 44\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 44\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Judgmental%20forecast%20adjustments%20over%20different%20time%20horizons&amp;journal=Omega&amp;doi=10.1016%2Fj.omega.2018.09.008&amp;volume=87&amp;pages=34-45&amp;publication_year=2019&amp;author=Broeke%2CM&amp;author=Baets%2CS&amp;author=Vereecke%2CA&amp;author=Baecke%2CP&amp;author=Vanderheyden%2CK\">                     Google Scholar<\/a>\u00a0                 <br \/>Arvan, M., Fahimnia, B., Reisi, M. &amp; Siemsen, E. Integrating human judgement into quantitative forecasting methods: A review. <i>Omega<\/i> <b>86<\/b>, 237\u2013252. <a href=\"https:\/\/doi.org\/10.1016\/j.omega.2018.07.012\">https:\/\/doi.org\/10.1016\/j.omega.2018.07.012<\/a> (2019).<br \/><a data-track=\"click\" rel=\"nofollow noopener\" data-track-label=\"10.1016\/j.omega.2018.07.012\" data-track-action=\"article reference\" href=\"https:\/\/doi.org\/10.1016%2Fj.omega.2018.07.012\" aria-label=\"Article reference 45\">Article<\/a>\u00a0     <a data-track=\"click\" data-track-action=\"google scholar reference\" data-track-label=\"link\" rel=\"nofollow noopener\" aria-label=\"Google Scholar reference 45\" href=\"http:\/\/scholar.google.com\/scholar_lookup?&amp;title=Integrating%20human%20judgement%20into%20quantitative%20forecasting%20methods%3A%20A%20review&amp;journal=Omega&amp;doi=10.1016%2Fj.omega.2018.07.012&amp;volume=86&amp;pages=237-252&amp;publication_year=2019&amp;author=Arvan%2CM&amp;author=Fahimnia%2CB&amp;author=Reisi%2CM&amp;author=Siemsen%2CE\">                     Google Scholar<\/a>\u00a0                 <br \/><a data-track=\"click\" data-track-action=\"download citation references\" data-track-label=\"link\" href=\"https:\/\/citation-needed.springer.com\/v2\/references\/10.1038\/s41598-022-20015-y?format=refman&amp;flavour=references\">Download references<svg width=\"16\" height=\"16\" focusable=\"false\" role=\"img\" aria-hidden=\"true\" class=\"u-icon\"><use xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" xlink:href=\"#global-icon-download\"><\/use><\/svg><\/a><br \/>The first author would like to thank Academic Leader Program (grant no. JXSGZL202203) for their financial support, thank Dr. Ting Yang and Dr. Qing Liu\u2019s advices, and thank Dr. Alexander Gegov and the School of Computing at University of Portsmouth who offered the invitation as a visiting researcher and shared the theory of Fuzzy Network. Thank two anonymous reviewers for the constructive comments and valuable advices.<br \/>College of Power Engineering, Naval University of Engineering, Wuhan, 430033, China<br \/>Xiaowei Wang,\u00a0Yanqiao Chen\u00a0&amp;\u00a0Jiashan Jin<br \/>Army Engineering University of PLA, Wuhan, 430075, China<br \/>Xiaowei Wang\u00a0&amp;\u00a0Baohua Zhang<br \/>You can also search for this author in                         <span class=\"c-article-identifiers\"><a class=\"c-article-identifiers__item\" href=\"http:\/\/www.ncbi.nlm.nih.gov\/entrez\/query.fcgi?cmd=search&amp;term=Xiaowei%20Wang\" data-track=\"click\" data-track-action=\"author link - pubmed\" data-track-label=\"link\" rel=\"nofollow\">PubMed<\/a><span class=\"u-hide\">\u00a0<\/span><a class=\"c-article-identifiers__item\" href=\"http:\/\/scholar.google.co.uk\/scholar?as_q=&amp;num=10&amp;btnG=Search+Scholar&amp;as_epq=&amp;as_oq=&amp;as_eq=&amp;as_occt=any&amp;as_sauthors=%22Xiaowei%20Wang%22&amp;as_publication=&amp;as_ylo=&amp;as_yhi=&amp;as_allsubj=all&amp;hl=en\" data-track=\"click\" data-track-action=\"author link - scholar\" data-track-label=\"link\" rel=\"nofollow\">Google Scholar<\/a><\/span><br \/>You can also search for this author in                         <span class=\"c-article-identifiers\"><a class=\"c-article-identifiers__item\" href=\"http:\/\/www.ncbi.nlm.nih.gov\/entrez\/query.fcgi?cmd=search&amp;term=Yanqiao%20Chen\" data-track=\"click\" data-track-action=\"author link - pubmed\" data-track-label=\"link\" rel=\"nofollow\">PubMed<\/a><span class=\"u-hide\">\u00a0<\/span><a class=\"c-article-identifiers__item\" href=\"http:\/\/scholar.google.co.uk\/scholar?as_q=&amp;num=10&amp;btnG=Search+Scholar&amp;as_epq=&amp;as_oq=&amp;as_eq=&amp;as_occt=any&amp;as_sauthors=%22Yanqiao%20Chen%22&amp;as_publication=&amp;as_ylo=&amp;as_yhi=&amp;as_allsubj=all&amp;hl=en\" data-track=\"click\" data-track-action=\"author link - scholar\" data-track-label=\"link\" rel=\"nofollow\">Google Scholar<\/a><\/span><br \/>You can also search for this author in                         <span class=\"c-article-identifiers\"><a class=\"c-article-identifiers__item\" href=\"http:\/\/www.ncbi.nlm.nih.gov\/entrez\/query.fcgi?cmd=search&amp;term=Jiashan%20Jin\" data-track=\"click\" data-track-action=\"author link - pubmed\" data-track-label=\"link\" rel=\"nofollow\">PubMed<\/a><span class=\"u-hide\">\u00a0<\/span><a class=\"c-article-identifiers__item\" href=\"http:\/\/scholar.google.co.uk\/scholar?as_q=&amp;num=10&amp;btnG=Search+Scholar&amp;as_epq=&amp;as_oq=&amp;as_eq=&amp;as_occt=any&amp;as_sauthors=%22Jiashan%20Jin%22&amp;as_publication=&amp;as_ylo=&amp;as_yhi=&amp;as_allsubj=all&amp;hl=en\" data-track=\"click\" data-track-action=\"author link - scholar\" data-track-label=\"link\" rel=\"nofollow\">Google Scholar<\/a><\/span><br \/>You can also search for this author in                         <span class=\"c-article-identifiers\"><a class=\"c-article-identifiers__item\" href=\"http:\/\/www.ncbi.nlm.nih.gov\/entrez\/query.fcgi?cmd=search&amp;term=Baohua%20Zhang\" data-track=\"click\" data-track-action=\"author link - pubmed\" data-track-label=\"link\" rel=\"nofollow\">PubMed<\/a><span class=\"u-hide\">\u00a0<\/span><a class=\"c-article-identifiers__item\" href=\"http:\/\/scholar.google.co.uk\/scholar?as_q=&amp;num=10&amp;btnG=Search+Scholar&amp;as_epq=&amp;as_oq=&amp;as_eq=&amp;as_occt=any&amp;as_sauthors=%22Baohua%20Zhang%22&amp;as_publication=&amp;as_ylo=&amp;as_yhi=&amp;as_allsubj=all&amp;hl=en\" data-track=\"click\" data-track-action=\"author link - scholar\" data-track-label=\"link\" rel=\"nofollow\">Google Scholar<\/a><\/span><br \/>X.W. and Y.C. wrote the main manuscript text, and J.J. and B.Z. implemented the case study. All authors reviewed the manuscript.<br \/>Correspondence to                 <a id=\"corresp-c1\" href=\"mailto:yannqiao_chen@163.com\">Yanqiao Chen<\/a>.<br \/>The authors declare no competing interests.<br \/>Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.<br \/><b>Open Access<\/b>  This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article&#8217;s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article&#8217;s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit <a href=\"http:\/\/creativecommons.org\/licenses\/by\/4.0\/\" rel=\"license\">http:\/\/creativecommons.org\/licenses\/by\/4.0\/<\/a>.<br \/><a data-track=\"click\" data-track-action=\"view rights and permissions\" data-track-label=\"link\" href=\"https:\/\/s100.copyright.com\/AppDispatchServlet?title=Fuzzy-clustering%20and%20fuzzy%20network%20based%20interpretable%20fuzzy%20model%20for%20prediction&amp;author=Xiaowei%20Wang%20et%20al&amp;contentID=10.1038%2Fs41598-022-20015-y&amp;copyright=The%20Author%28s%29&amp;publication=2045-2322&amp;publicationDate=2022-09-29&amp;publisherName=SpringerNature&amp;orderBeanReset=true&amp;oa=CC%20BY\">Reprints and Permissions<\/a><br \/>Wang, X., Chen, Y., Jin, J. <i>et al.<\/i> Fuzzy-clustering and fuzzy network based interpretable fuzzy model for prediction.                     <i>Sci Rep<\/i> <b>12<\/b>, 16279 (2022). https:\/\/doi.org\/10.1038\/s41598-022-20015-y<br \/><a data-test=\"citation-link\" data-track=\"click\" data-track-action=\"download article citation\" data-track-label=\"link\" data-track-external=\"\" href=\"https:\/\/citation-needed.springer.com\/v2\/references\/10.1038\/s41598-022-20015-y?format=refman&amp;flavour=citation\">Download citation<svg width=\"16\" height=\"16\" focusable=\"false\" role=\"img\" aria-hidden=\"true\" class=\"u-icon\"><use xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" xlink:href=\"#global-icon-download\"><\/use><\/svg><\/a><br \/>Received<span class=\"u-hide\">: <\/span><span class=\"c-bibliographic-information__value\"><time datetime=\"2022-06-17\">17 June 2022<\/time><\/span><br \/>Accepted<span class=\"u-hide\">: <\/span><span class=\"c-bibliographic-information__value\"><time datetime=\"2022-09-07\">07 September 2022<\/time><\/span><br \/>Published<span class=\"u-hide\">: <\/span><span class=\"c-bibliographic-information__value\"><time datetime=\"2022-09-29\">29 September 2022<\/time><\/span><br \/><abbr title=\"Digital Object Identifier\">DOI<\/abbr><span class=\"u-hide\">: <\/span><span class=\"c-bibliographic-information__value\">https:\/\/doi.org\/10.1038\/s41598-022-20015-y<\/span><br \/>Anyone you share the following link with will be able to read this content:<br \/>Sorry, a shareable link is not currently available for this article.<\/p>\n<p>                             Provided by the Springer Nature SharedIt content-sharing initiative                         <br \/>By submitting a comment you agree to abide by our <a href=\"\/info\/tandc.html\">Terms<\/a> and <a href=\"\/info\/community-guidelines.html\">Community Guidelines<\/a>. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.<br \/>Advertisement<br \/>                 <span class=\"c-meta__item\">                     Scientific Reports (<i>Sci Rep<\/i>)                 <\/span>                           <span class=\"c-meta__item\">         <abbr title=\"International Standard Serial Number\">ISSN<\/abbr> <span itemprop=\"onlineIssn\">2045-2322<\/span> (online)     <\/span>                                           <br \/>&copy; 2022 Springer Nature Limited<br \/>Sign up for the <em>Nature Briefing<\/em> newsletter \u2014 what matters in science, free to your inbox daily.<\/p>\n<p><a href=\"https:\/\/www.nature.com\/articles\/s41598-022-20015-y\">source<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.Advertisement [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":72713,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-72712","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/loopbeatzz.info\/wp\/wp-json\/wp\/v2\/posts\/72712","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/loopbeatzz.info\/wp\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/loopbeatzz.info\/wp\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/loopbeatzz.info\/wp\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/loopbeatzz.info\/wp\/wp-json\/wp\/v2\/comments?post=72712"}],"version-history":[{"count":0,"href":"https:\/\/loopbeatzz.info\/wp\/wp-json\/wp\/v2\/posts\/72712\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/loopbeatzz.info\/wp\/wp-json\/wp\/v2\/media\/72713"}],"wp:attachment":[{"href":"https:\/\/loopbeatzz.info\/wp\/wp-json\/wp\/v2\/media?parent=72712"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/loopbeatzz.info\/wp\/wp-json\/wp\/v2\/categories?post=72712"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/loopbeatzz.info\/wp\/wp-json\/wp\/v2\/tags?post=72712"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}