CMA-ES with Two-Point Step-Size Adaptation - Archive ouverte HAL Access content directly
Reports (Research Report) Year : 2008

CMA-ES with Two-Point Step-Size Adaptation

Abstract

We combine a refined version of two-point step-size adaptation with the covariance matrix adaptation evolution strategy (CMA-ES). Additionally, we suggest polished formulae for the learning rate of the covariance matrix and the recombination weights. In contrast to cumulative step-size adaptation or to the 1/5-th success rule, the refined two-point adaptation (TPA) does not rely on any internal model of optimality. In contrast to conventional self-adaptation, the TPA will achieve a better target step-size in particular with large populations. The disadvantage of TPA is that it relies on two additional objective function evaluations.
Fichier principal
Vignette du fichier
twopointSA.pdf (169.26 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

inria-00276854 , version 1 (02-05-2008)
inria-00276854 , version 2 (02-05-2008)
inria-00276854 , version 3 (03-05-2008)
inria-00276854 , version 4 (13-05-2008)
inria-00276854 , version 5 (18-05-2008)

Identifiers

  • HAL Id : inria-00276854 , version 5
  • ARXIV : 0805.0231

Cite

Nikolaus Hansen. CMA-ES with Two-Point Step-Size Adaptation. [Research Report] RR-6527, INRIA. 2008. ⟨inria-00276854v5⟩
419 View
396 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More