Influence of agent-α statistics on cluster-size distribution

Abstract

Empirical cluster sizes have been found to exhibit scaling properties. A behavioral agent-based mechanism has been proposed to explain how the scaling emerges from simple local rules. This mechanism is based on the idea that social entities such as animals and human beings aggregate together by means of behavioral mimicry, technically known as allelomimesis. In the agent-based model, the tendency for an agent to behave allelomimetically is represented by a probability α. It has been shown that α is capable of tuning the scaling properties of the cluster-size statistics. It was assumed in previous studies that all agents in a particular population have the same value of α. In this study, we consider that α varies among agents of a population. A generalized bimodal distribution for α – a superposition of two Gaussian distributions – is assumed. We investigate the effects of modifying this statistical distribution on the scaling properties of the size distribution of simulated clusters.