Application of the Shuffled Frog Leaping Algorithm (SFLA) in Constructing Fuzzy Classification Systems

Abstract

A Fuzzy Inference System (FIS) is a way of mapping an input space to an output space using the fuzzy logic. FISs are widely used to solve classification problems. The Shuffled Frog Leaping Algorithm (SFLA) is a metaheuristic inspired by the natural evolution of frogs in searching for the largest source of food. By using local and global searches simultaneously, SFLA is effective in solving various optimization problems. This paper first proposes a new method to create zero-order Sugeno Fuzzy Inference Systems using SFLA. Then, the paper introduces an approach to use resulting SFLA-based Fuzzy Inference Systems to build test oracles. Test oracle is a mechanism for determining whether a test on a software program has passed or failed. The experimental results show that SFLA creates fuzzy systems more efficiently than three other evolutionary algorithms, including Genetic Algorithm (GA), Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO). Moreover, with respect to the accuracy and convergence speed criteria, SFLA and PSO outperform other evolutionary algorithms, while their performances are comparable to each other. At last, the experimental results indicate that SFLA-based FISs can be used to create test oracles with acceptable accuracy.