English

Runtime Analysis of Evolutionary Algorithms via Symmetry Arguments

Neural and Evolutionary Computing 2021-09-21 v3 Data Structures and Algorithms

Abstract

We use an elementary argument building on group actions to prove that the selection-free steady state genetic algorithm analyzed by Sutton and Witt (GECCO 2019) takes an expected number of Ω(2n/n)\Omega(2^n / \sqrt n) iterations to find any particular target search point. This bound is valid for all population sizes μ\mu. Our result improves over the previous lower bound of Ω(exp(nδ/2))\Omega(\exp(n^{\delta/2})) valid for population sizes μ=O(n1/2δ)\mu = O(n^{1/2 - \delta}), 0<δ<1/20 < \delta < 1/2.

Keywords

Cite

@article{arxiv.2006.04663,
  title  = {Runtime Analysis of Evolutionary Algorithms via Symmetry Arguments},
  author = {Benjamin Doerr},
  journal= {arXiv preprint arXiv:2006.04663},
  year   = {2021}
}

Comments

Minor changes compared to the previous version

R2 v1 2026-06-23T16:08:58.976Z