Title :
Learning by simulating evolution in automatic fuzzy systems synthesis
Author :
Buhusi, Catalin V.
Author_Institution :
Inst. of Comput. Sci., Romanian Acad., Iasi, Romania
Abstract :
In this paper we present a genetic learning method for the automatic synthesis of a class of fuzzy systems with variable number of rules of specific form, namely dynamic self-organizing fuzzy systems (DSOFSs). The goal of the genetic synthesis is the search of an optimal set of such rules when synthesizing a fuzzy system for a specific problem. This optimal set of rules must reach for some desired features of the fuzzy system (such as minimal number of rules etc.) In order to reach this goal the proposed genetic learning method uses an unequal crossover operator which allows the synthesis of the variable structure of the fuzzy system. The results of the genetic synthesis in a pattern recognition problem are presented. The results emphasize the impact of the specific genetic learning method over the desired characteristics of the fuzzy system
Keywords :
adaptive systems; fuzzy systems; genetic algorithms; learning (artificial intelligence); self-adjusting systems; variable structure systems; automatic fuzzy systems synthesis; dynamic self-organizing fuzzy systems; evolution simulation; genetic learning method; pattern recognition; unequal crossover operator; variable structure synthesis; Artificial intelligence; Computational modeling; Computer science; Fuzzy systems; Genetic algorithms; Learning systems; Network synthesis; Organizing; Pattern recognition; Proposals;
Conference_Titel :
Fuzzy Systems, 1994. IEEE World Congress on Computational Intelligence., Proceedings of the Third IEEE Conference on
Conference_Location :
Orlando, FL
Print_ISBN :
0-7803-1896-X
DOI :
10.1109/FUZZY.1994.343632
