Unbounded Population MO-CMA-ES for the Bi-Objective BBOB Test Suite

Oswin Krause 1 Tobias Glasmachers 2 Nikolaus Hansen 3 Christian Igel 1
3 TAO - Machine Learning and Optimisation
LRI - Laboratoire de Recherche en Informatique, UP11 - Université Paris-Sud - Paris 11, Inria Saclay - Ile de France, CNRS - Centre National de la Recherche Scientifique : UMR8623
Abstract : The unbounded population multi-objective covariance matrix adaptation evolution strategy (UP-MO-CMA-ES) aims at maximizing the total hypervolume covered by all evaluated points. It adds all non-dominated solutions found to its population and employs Gaussian mutations with adaptive covariance matrices to also solve ill-conditioned problems. A novel recombination operator adapts the covariance matrices to point along the Pareto front. The UP-MO-CMA-ES is combined with a parallel exploration strategy and empirically evaluated on the bi-objective BBOB-biobj benchmark problems. Results show that the algorithm can reliably solve ill-conditioned problems as well as weakly-structured problems. However, it is less suited for the rugged multi-modal objective functions in the benchmark.
Keywords : CMA-ES MO-CMA-ES
Complete list of metadatas

Contributor : Nikolaus Hansen <>
Submitted on : Friday, October 14, 2016 - 3:03:11 PM
Last modification on : Friday, March 22, 2019 - 4:46:03 PM



Oswin Krause, Tobias Glasmachers, Nikolaus Hansen, Christian Igel. Unbounded Population MO-CMA-ES for the Bi-Objective BBOB Test Suite. GECCO'16 - Companion of Proceedings of the 2016 Genetic and Evolutionary Computation Conference, ACM, Jul 2016, Denver, United States. pp.1177-1184, ⟨10.1145/2908961.2931699⟩. ⟨hal-01381653⟩



Record views