Two enhancements for improving the convergence speed of a robust multi-objective coevolutionary algorithm.

Zavoianu, Alexandru-Ciprian; Saminger-Platz, Susanne; Lughofer, Edwin; Amrhein, Wolfgang

Abstract

We describe two enhancements that significantly improve the rapid convergence behavior of DECM02 - a previously proposed robust coevolutionary algorithm that integrates three different multi-objective space exploration paradigms: differential evolution, two-tier Pareto-based selection for survival and decomposition-based evolutionary guidance. The first enhancement is a refined active search adaptation mechanism that relies on run-time sub-population performance indicators to estimate the convergence stage and dynamically adjust and steer certain parts of the coevolutionary process in order to improve its overall efficiency. The second enhancement consists in a directional intensification operator that is applied in the early part of the run during the decomposition-based search phases. This operator creates new random local linear individuals based on the recent historically successful solution candidates of a given directional decomposition vector. As the two efficiency-related enhancements are complementary, our results show that the resulting coevolutionary algorithm is a highly competitive improvement of the baseline strategy when considering a comprehensive test set aggregated from 25 (standard) benchmark multi-objective optimization problems.

Citation

ZAVOIANU, A.-C., SAMINGER-PLATZ, S., LUGHOFER, E. and AMRHEIN, W. 2018. Two enhancements for improving the convergence speed of a robust multi-objective coevolutionary algorithm. In Aguirre, H. (ed.) Proceedings of the 2018 Genetic and evolutionary computation conference (GECCO'18), 15-19 July 2018, Kyoto, Japan. New York: Association for Computing Machinery [online], pages 793-800. Available from: https://doi.org/10.1145/3205455.3205549

• Conference Name: 2018 Genetic and evolutionary computation conference (GECCO'18)
• Conference Location: Kyoto, Japan
• Start Date: Jul 15, 2018
• End Date: Jul 19, 2018
• Acceptance Date: Apr 17, 2018
• Online Publication Date: Jul 2, 2018
• Publication Date: Jul 2, 2018
• Deposit Date: Mar 2, 2020
• Publicly Available Date: Apr 2, 2020
• Publisher: Association for Computing Machinery (ACM)
• Pages: 793-800
• ISBN: 9781450356183
• DOI: https://doi.org/10.1145/3205455.3205549
• Keywords: Coevolution; Sub-population dynamics; Differential evolution; Run-time performance assessment; Multi-objective optimisation
• Public URL: https://rgu-repository.worktribe.com/output/870387