Fuzzy system modeling with fuzzy functions

In this study, new fuzzy system modeling approaches are proposed. Performances of proposed methods and improved fuzzy functions with improved fuzzy clustering approach proposed by Türkşen [2008] and Çelikyılmaz and Türkşen [2007, 2008, 2009] which refers to this study are analyzed by experiments. With this study, two new inference algorithms are proposed to model the systems to fuzzy c-regression model proposed by Hataway and Bezdek [10] and fuzzy model proposed by Höppner and Klawonn [2003] algorithms. In addition to this, an alternative clustering algorithm (Alternative Fuzzy Model - AFM) is proposed by improving the fuzzy model of Höppner and Klawonn [2003]. According to this, an inference algorithm is proposed for AFM to model the system. Based on the proposed improvements newly proposed model results are compared with the previously proposed models.