Evolving agent-based models using self-adaptive complexification

•We examine evolutionary optimization of complex systems.•Evolutionary optimization is more difficult the more dimensions the problem has.•Complexification is useful to improve search strategy of evolutionary algorithms.•We show that self adapting complexification further improves the results.

This paper focuses on parameter search for multi-agent based models using evolutionary algorithms. Large numbers and variable dimensions of parameters require an optimization method which can efficiently handle a high dimensional search space. We are proposing the use of complexification as it emulates the natural way of evolution by starting with a small constrained search space and expanding it as the evolution progresses. To further improve performance we suggest and experiment with methods of self-adaptation to enable the algorithm to adjust its parameters individually to the problem at hand. We examined the effects of these methods on an EA by evolving parameters for two multi-agent based models.