Performance and Population Diversity of Steady-State GA on Fitness Landscapes with Neutral Networks

Abstract

Neutral networks, which are landscapes containing neighboring points of equal fitness, have attracted much research interest in recent years.In this paper, we applied steady-state GAs with different deletion methods to two test functions in order to investigate the relationship between the GA performance and the population diversity.The results show that a population conversed in fitness space moves faster on a neutral nework, but can easily get trapped on neutral networks which do not lead to neutral networks of higher fitness as ruggedness increases.