Maximum Likelihood-based Online Adaptation of Hyper-parameters in CMA-ES - Archive ouverte HAL Access content directly
Conference Papers Year : 2014

Maximum Likelihood-based Online Adaptation of Hyper-parameters in CMA-ES

(1) , (2, 3) , (4) , (3)
1
2
3
4

Abstract

The Covariance Matrix Adaptation Evolution Strategy (CMA-ES) is widely accepted as a robust derivative-free continuous optimization algorithm for non-linear and non-convex optimization problems. CMA-ES is well known to be almost parameterless, meaning that only one hyper-parameter, the population size, is proposed to be tuned by the user. In this paper, we propose a principled approach called self-CMA-ES to achieve the online adaptation of CMA-ES hyper-parameters in order to improve its overall performance. Experimental results show that for larger-than-default population size, the default settings of hyper-parameters of CMA-ES are far from being optimal, and that self-CMA-ES allows for dynamically approaching optimal settings.
Fichier principal
Vignette du fichier
selfCMA.pdf (991.45 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01003504 , version 1 (10-06-2014)
hal-01003504 , version 2 (11-06-2014)

Identifiers

Cite

Ilya Loshchilov, Marc Schoenauer, Michèle Sebag, Nikolaus Hansen. Maximum Likelihood-based Online Adaptation of Hyper-parameters in CMA-ES. 13th International Conference on Parallel Problem Solving from Nature (PPSN 2014), Sep 2014, Ljubljana, Slovenia. pp.70-79. ⟨hal-01003504v2⟩
483 View
354 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More