Is CC-(1+1) EA more efficient than (1+1) EA on either separable or inseparable problems?

Per Kristian Lehre, Shishen Lin

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper is about the cooperative co-evolutionary algorithms. This paper investigates cooperative co-evolutionary algorithms (CoEAs) for large-scale optimization problems, focusing on the runtime analysis to understand their behavior. By proving that the basic cooperative co-evolutionary (1+1) EA has an expected optimization time of Θ(n log n) on linear functions, it solves an open conjecture. Empirical analysis on NK-LANDSCAPE and k-MAXSAT problems shows that performance can be optimized by adjusting block length, providing more precise runtime bounds and insights on more complicated inseparable problems.
Original languageEnglish
Title of host publication IEEE Congress on Evolutionary Computation
PublisherIEEE
Number of pages9
ISBN (Electronic)9798350314588
ISBN (Print)9798350314595
DOIs
Publication statusPublished - 25 Sept 2023
EventIEEE 2023 Congress on Evolutionary Computation - Swissotel Chicago, Chicago, United States
Duration: 1 Jul 20235 Jul 2023

Conference

ConferenceIEEE 2023 Congress on Evolutionary Computation
Abbreviated titleCEC 2023
Country/TerritoryUnited States
CityChicago
Period1/07/235/07/23

Keywords

  • coevolution

Fingerprint

Dive into the research topics of 'Is CC-(1+1) EA more efficient than (1+1) EA on either separable or inseparable problems?'. Together they form a unique fingerprint.

Cite this