Competitive evaluation of selected evolutionary algorithms and SOMA

Abstract

In this paper three evolutionary algorithms are compared, in particular, of the Self-Organizing Migration Algorithm (SOMA) as a most important one is put into the contrast with Differential Evolution (DE) and Particle Swarm Optimization (PSO). In order to compare performances of the above-mentioned algorithms, selected objective functions have been tested. In total 15 different benchmark functions were used and each considered algorithm was employed 100 times on each one producing 4500 set experimental runs. Acquired results were then statistically evaluated and compared. The paper also describes individual parameters, strategies, and some of the termination criteria of the algorithms.