An evolutionary and neighborhood-based algorithm for optimization under low budget requirements

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

Abstract

Noisy optimization problems pose several challenges to optimization algorithms under limited computational resources. The algorithm must balance the need to explore the search space and to exploit promising regions of this space. In this work we describe an algorithm that combines previous ideas to tackle the GECCO 2021 industrial challenge using an exploration and an exploitation step. The first step consists in an evolutionary algorithm combined with an experimental design, while the second phase is a neighborhood search embedded within a multi-armed greedy approach. The resulting algorithm is not only applicable to the challenge but also to more general problems with varying conditions.

Original languageEnglish
Title of host publicationGECCO 2021 Companion - Proceedings of the 2021 Genetic and Evolutionary Computation Conference Companion
PublisherAssociation for Computing Machinery, Inc
Pages17-18
Number of pages2
ISBN (Electronic)9781450383516
DOIs
StatePublished - 7 Jul 2021
Externally publishedYes
Event2021 Genetic and Evolutionary Computation Conference, GECCO 2021 - Virtual, Online, France
Duration: 10 Jul 202114 Jul 2021

Publication series

NameGECCO 2021 Companion - Proceedings of the 2021 Genetic and Evolutionary Computation Conference Companion

Conference

Conference2021 Genetic and Evolutionary Computation Conference, GECCO 2021
Country/TerritoryFrance
CityVirtual, Online
Period10/07/2114/07/21

Keywords

  • design of experiments
  • estimation of distribution algorithms
  • greedy
  • optimization of simulations

Fingerprint

Dive into the research topics of 'An evolutionary and neighborhood-based algorithm for optimization under low budget requirements'. Together they form a unique fingerprint.

Cite this