Skip to Main content Skip to Navigation
Conference papers

DCMA, yet another derandomization in covariance-matrix-adaptation

Olivier Teytaud 1 Sylvain Gelly 1 
1 TANC - Algorithmic number theory for cryptology
LIX - Laboratoire d'informatique de l'École polytechnique [Palaiseau], Inria Saclay - Ile de France
Abstract : In a preliminary part of this paper, we analyze the neces- sity of randomness in evolution strategies. We conclude to the necessity of ”continuous”-randomness, but with a much more limited use of randomness than what is commonly used in evolution strategies. We then apply these results to CMA- ES, a famous evolution strategy already based on the idea of derandomization, which uses random independent Gaus- sian mutations. We here replace these random independent Gaussian mutations by a quasi-random sample. The mod- ification is very easy to do, the modified algorithm is com- putationally more efficient and its convergence is faster in terms of the number of iterates for a given precision.
Document type :
Conference papers
Complete list of metadata

Cited literature [22 references]  Display  Hide  Download
Contributor : Olivier Teytaud Connect in order to contact the contributor
Submitted on : Wednesday, September 19, 2007 - 3:46:02 PM
Last modification on : Friday, February 4, 2022 - 3:19:22 AM
Long-term archiving on: : Friday, April 9, 2010 - 2:28:41 AM


Files produced by the author(s)


  • HAL Id : inria-00173207, version 1



Olivier Teytaud, Sylvain Gelly. DCMA, yet another derandomization in covariance-matrix-adaptation. GECCO, 2007, London, United Kingdom. ⟨inria-00173207⟩



Record views


Files downloads