Abstract
In this study, designing networks is formulated as an optimization problem with objective functions regarding characteristics of networks. We introduce an approach to find the contributing factors to construct complex networks by comparing and examining the characteristics between generated networks and the complex network. In this paper, we apply a genetic algorithm (GA) as optimization method and examine the network design focusing on characteristics of the complex network as basic studies for our proposal approach. Here the average shortest movement distance and/or the cluster coefficient are used as objective functions to construct networks. Through the experiments, it was clear that there was trade-off between these two factors and the multi-objective design was more appropriate than single-objective one, for applying our approach to real networks.
