Diversity maintenance perspective: An analysis of exploratory power and function optimization in the context of adaptive genetic algorithms

Abstract

In order to increase the probability of finding optimal solution, GAs must maintain a balance between the exploration and exploitation. Maintaining population diversity not only prevents premature convergence but also provides a better coverage of the search space. Diversity measures are traditionally used to analyze evolutionary algorithms rather than guiding them. This chapter discusses the applicability of updation phase of binary trie coding scheme [BTCS] in introducing as well as maintaining population diversity. Here, the robustness of BTCS is compared with informed hybrid adaptive genetic algorithm (IHAGA), which works by adaptively changing the probabilities of crossover and mutation based on the fitness results of the respective offsprings in the next generation.