Sudholt, D. orcid.org/0000-0001-6020-1646 (2018) On the Robustness of Evolutionary Algorithms to Noise: Refined Results and an Example Where Noise Helps. In: Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2018). Genetic and Evolutionary Computation Conference (GECCO 2018), 15-19 Jul 2018, Kyoto, Japan. ACM ISBN 978-1-4503-5618-3
Abstract
We present reined results for the expected optimisation time of the (1+1) EA and the (1+λ) EA on LeadingOnes in the prior noise model, where in each itness evaluation the search point is altered before evaluation with probability p. Previous work showed that the (1+1) EA runs in polynomial time if p = O((logn)/n 2 ) and needs superpolynomial time if p = Ω((logn)/n), leaving a huge gap for which no results were known. We close this gap by showing that the expected optimisation time is Θ(n 2 ) · exp(Θ(pn2 )), allowing for the irst time to locate the threshold between polynomial and superpolynomial expected times at p = Θ((logn)/n 2 ). Hence the (1+1) EA on LeadingOnes is much more sensitive to noise than previously thought. We also show that ofspring populations of size λ ≥ 3.42 logn can efectively deal with much higher noise than known before. Finally, we present an example of a rugged landscape where prior noise can help to escape from local optima by blurring the landscape and allowing a hill climber to see the underlying gradient.
Metadata
Item Type: | Proceedings Paper |
---|---|
Authors/Creators: |
|
Copyright, Publisher and Additional Information: | © 2018 ACM. This is an author-produced version of a paper accepted for publication. Uploaded in accordance with the publisher's self-archiving policy. |
Keywords: | Evolutionary algorithms; noisy optimisation; runtime analysis; theory |
Dates: |
|
Institution: | The University of Sheffield |
Academic Units: | The University of Sheffield > Faculty of Engineering (Sheffield) > Department of Computer Science (Sheffield) |
Depositing User: | Symplectic Sheffield |
Date Deposited: | 16 May 2018 13:36 |
Last Modified: | 19 Dec 2022 13:49 |
Published Version: | https://doi.org/10.1145/3205455.3205595 |
Status: | Published online |
Publisher: | ACM |
Refereed: | Yes |
Identification Number: | 10.1145/3205455.3205595 |
Open Archives Initiative ID (OAI ID): | oai:eprints.whiterose.ac.uk:130702 |