PSO-GAScript: A Domain-specific Scripting Language for Meta-heuristics Algorithm

Abstract

PSO-GAScript is a domain-specific scripting language designed to support easy and rapid implementation of meta-heuristics algorithms focused on Particle Swarm Optimization (PSO) and Genetic Algorithm (GA). The programming language has been developed to allow the hybridization of the two meta-heuristics algorithms. Hybridizations between PSO and GA are proven to be a comprehensive tool for solving different kinds of optimization problems. Moreover, the two algorithms have achieved a remarkable improvement from the adaptation of dynamic parameterization. Nevertheless, implementing the suitable hybrid algorithms is a considerably difficult, which in most cases is time consuming. To the best of our knowledge, the existing tools are not adequately designed to enable users to easily develop the meta-heuristics hybridization of PSO-GA with dynamic parameterizations. This paper presents the fundamental research methodology of domain-specific scripting language from the language design, constructs and evaluations focused on the case of PSO-GA hybridization and dynamic parameterizations. The PSO-GAScript are shown to easily use with minimal number of codes lines and concisely describe the meta-heuristics algorithms in a directly publishable form.