Benchmarking a weighted negative covariance matrix update on the BBOB-2010 noiseless testbed

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

We implement a weighted negative update of the covariance matrix in the CMA-ES - weighted active CMA-ES or, in short, aCMA-ES. We benchmark the IPOP-aCMA-ES and compare the performance with the IPOP-CMA-ES on the BBOB-2010 noiseless testbed in dimensions between 2 and 40. On nine out of 12 essentially unimodal functions, the aCMA is faster than CMA, in particular in larger dimension. On at least three functions it also leads to a (slightly) better scaling with the dimension. In none of the 24 benchmark functions aCMA appears to be significantly worse in any dimension. On two and five functions, IPOP-CMA-ES and IPOP-aCMA-ES respectively exceed the record observed during BBOB-2009.

Original languageEnglish
Title of host publicationProceedings of the 12th Annual Genetic and Evolutionary Computation Conference, GECCO '10 - Companion Publication
Pages1673-1680
Number of pages8
DOIs
Publication statusPublished - 30 Aug 2010
Externally publishedYes
Event12th Annual Genetic and Evolutionary Computation Conference, GECCO-2010 - Portland, OR, United States
Duration: 7 Jul 201011 Jul 2010

Publication series

NameProceedings of the 12th Annual Genetic and Evolutionary Computation Conference, GECCO '10 - Companion Publication

Conference

Conference12th Annual Genetic and Evolutionary Computation Conference, GECCO-2010
Country/TerritoryUnited States
CityPortland, OR
Period7/07/1011/07/10

Keywords

  • Active CMA-ES
  • Benchmarking
  • Black-box optimization
  • CMA-ES
  • IPOP-CMA-ES

Fingerprint

Dive into the research topics of 'Benchmarking a weighted negative covariance matrix update on the BBOB-2010 noiseless testbed'. Together they form a unique fingerprint.

Cite this