Genetic Algorithm Based Fuzzy System Optimization for Breast Cancer Detection

Abstract

Fuzzy logic is a branch of mathematics that focuses on the partial truth concept and approximation. The reasoning system of the theory has strong quantitative performance (accuracy) as well as language representation (interpretability). Designing a fuzzy system, on the other hand, is a difficult undertaking that necessitates the identification of numerous parameters. The performance of a fuzzy system is determined by several factors, including the fuzzy rules that are used and the membership functions. This function is responsible for mapping crisp inputs of the system to linguistic variables. In general designing these parameters is considered a difficult undertaking that necessitates the assistance of a fuzzy system expert or an effective optimization procedure. In order to overcome this problem, a genetic approach was used in this paper to create efficient fuzzy systems. The Wisconsin breast cancer diagnostic (WBCD) problem is then subjected to this optimization technique. This strategy was effective in developing fuzzy based systems which uses a smaller number of rules and provide a higher performance of over 93%.