ABSTRACT

Genetic fuzzy classification system is a hot spot of fuzzy systems, but the contradiction of its accuracy and interpretability is a problem. Genetic fuzzy classification system is a hybridization of fuzzy system and genetic algorithm, which has become a hot spot of fuzzy system. During the process of mutation, parameters based on computing with words are used to adjust the shape of the membership function. Simulation experiments of Iris data set show that the new method has higher classification accuracy. Professor Zadeh considered that adding language qualifiers can make a connection between data information and computing with words. Language modifiers are added to improve the system accuracy, which reduces its convergence rate. Expert selection is used in initializing process to improve the convergence speed. According to computing with words, four language qualifiers were added during the mutate process to modify the shape of the membership function. These are “extreme”, “very”, “more or less”, and “little”.