On the transforming of the indices selection mechanism inside differential evolution into complex network

Abstract

This research deals with complex networks framework for evolutionary algorithms. This paper aims on the experimental investigations on the time development and influence of different randomization types, different strategies for Differential Evolution (DE) through the analysis of complex network as a record of population dynamics and indices selection. The population is visualized as an evolving complex network, which exhibits non-trivial features such as adjacency graph, centralities, clustering coefficient and other attributes showing efficiency of the network. Experiments were performed for different DE strategies, several different randomization types and simple test function.