A Novel Evolutionary-Fuzzy System for Function Approximation and Comparison the Robust of Used Evolutionary Algorithms
Source: Islamic Azad University
Despite the previous successful history of fuzzy systems, the lack of learning capabilities characterizing most of the works generated a certain interest to design a fuzzy rule base with added learning capabilities. Evolutionary Algorithms (EAs) have inspired new design for fuzzy systems with added learning authorities. The purpose of the present study is to compare the capability of four evolutionary algorithms including Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Differential Evolution (DE) and Bees Algorithm (BA) to improve the capability of a new strength fuzzy inference system for nonlinear function approximation.