Maximum Likelihood-Based Online Adaptation of Hyper-Parameters in CMA-ES

Ilya Loshchilov¹, Marc Schoenauer^2,3, Michèle Sebag^3,2, and Nikolaus Hansen^2,3

¹Laboratory of Intelligent Systems, École Polytechnique Fédérale de Lausanne, Switzerland
Ilya.Loshchilov@epfl.ch

²TAO Project-team, INRIA Saclay, Île-de-France, France

³Laboratoire de Recherche en Informatique (UMR CNRS 8623), Université Paris-Sud, 91128, Orsay Cedex, France
FirstName.LastName@inria.fr

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.

LNCS 8672, p. 70 ff.

Full article in PDF | BibTeX

lncs@springer.com
© Springer International Publishing Switzerland 2014