Hypervolume-Based Surrogate Model for MO-CMA-ES

0459141 - ÚI 2017 RIV US eng C - Conference Paper (international conference)
Pilát, M. - Neruda, Roman
Hypervolume-Based Surrogate Model for MO-CMA-ES.
Proceedings of the 27th IEEE International Conference on Tools with Artificial Intelligence. Piscataway: IEEE Computer Society, 2015, s. 604-611. ISBN 978-1-5090-0163-7. ISSN 1082-3409.
[ICTAI 2015. IEEE International Conference on Tools with Artificial Intelligence /27./. Vietri sul Mare (IT), 09.11.2015-11.11.2015]
R&D Projects: GA ČR GA15-19877S
Institutional support: RVO:67985807
Keywords : MO-CMA-ES * hypervolume * surrogate modelling * multi-objective optimization
Subject RIV: IN - Informatics, Computer Science

Evolutionary algorithms are among the best multi-objective optimizers, but the large number of objective function evaluations they require makes it hard to use them to solve certain real-life tasks. In this work we present a surrogate-based local search for the multi-objective covariance matrix adaption evolution strategy (MO-CMA-ES). The local search is based on the estimation of hypervolume contribution of each individual and maximization of this contribution. This work extends our previous work and makes such surrogate models applicable to problems with more than two objectives. Moreover, it uses a unique feature of MO-CMA-ES to make the local search more effective. The results indicate that the algorithm can find solutions of the same quality as MO-CMA-ES while using 30-50 percent less objective function evaluations.
Permanent Link: http://hdl.handle.net/11104/0259383