Explaining SOMA: The relation of stochastic perturbation to population diversity and parameter space coverage

Abstract

The Self-Organizing Migrating Algorithm (SOMA) is enjoying a renewed interest of the research community, following recent achievements in various application areas and renowned performance competitions. In this paper, we focus on the importance and effect of the perturbation operator in SOMA as the perturbation is one of the fundamental inner principles of SOMA. In this in-depth study, we present data, visualizations, and analysis of the effect of the perturbation on the population, its diversity and average movement patterns. We provide evidence that there is a direct relation between the perturbation intensity (set by control parameter prt) and the rate of diversity loss. The perturbation setting further affects the exploratory ability of the algorithm, as is demonstrated here by analysing the parameter space coverage of the population. We aim to provide insight and explanation of the impact of perturbation in SOMA for future researchers and practitioners. © 2021 ACM.