Comparison of Performance between Different Selection Strategies on Simple Genetic Algorithms

Abstract

This paper presents the comparison of performance on a simple genetic algorithm (SGA) using roulette wheel selection and tournament selection. A SGA is mainly composed of three genetic operations, which are selection, crossover and mutation. With the same crossover and mutation operation, the simulation results are studied by comparing different selection strategies which are discussed in this paper. Qualitative analysis of the selection strategies is depicted, and the numerical experiments show that SGA with tournament selection strategy converges much faster than roulette wheel selection