English

Runtime Analysis for Self-adaptive Mutation Rates

Neural and Evolutionary Computing 2018-12-03 v1 Artificial Intelligence Data Structures and Algorithms Machine Learning

Abstract

We propose and analyze a self-adaptive version of the (1,λ)(1,\lambda) evolutionary algorithm in which the current mutation rate is part of the individual and thus also subject to mutation. A rigorous runtime analysis on the OneMax benchmark function reveals that a simple local mutation scheme for the rate leads to an expected optimization time (number of fitness evaluations) of O(nλ/logλ+nlogn)O(n\lambda/\log\lambda+n\log n) when λ\lambda is at least ClnnC \ln n for some constant C>0C > 0. For all values of λClnn\lambda \ge C \ln n, this performance is asymptotically best possible among all λ\lambda-parallel mutation-based unbiased black-box algorithms. Our result shows that self-adaptation in evolutionary computation can find complex optimal parameter settings on the fly. At the same time, it proves that a relatively complicated self-adjusting scheme for the mutation rate proposed by Doerr, Gie{\ss}en, Witt, and Yang~(GECCO~2017) can be replaced by our simple endogenous scheme. On the technical side, the paper contributes new tools for the analysis of two-dimensional drift processes arising in the analysis of dynamic parameter choices in EAs, including bounds on occupation probabilities in processes with non-constant drift.

Keywords

Cite

@article{arxiv.1811.12824,
  title  = {Runtime Analysis for Self-adaptive Mutation Rates},
  author = {Benjamin Doerr and Carsten Witt and Jing Yang},
  journal= {arXiv preprint arXiv:1811.12824},
  year   = {2018}
}
R2 v1 2026-06-23T06:27:05.881Z