Self-adaptation of mutation rates in non-elitist populations

The runtime of evolutionary algorithms (EAs) depends critically on their parameter settings, which are often problem-specific. Automated schemes for parameter tuning have been developed to alleviate the high costs of manual parameter tuning. Experimental results indicate that self-adaptation, where...

Full description

Bibliographic Details
Main Authors: Lehre, Per Kristian, Dang, Duc-Cuong
Format: Conference or Workshop Item
Published: 2016
Online Access:https://eprints.nottingham.ac.uk/34365/
_version_ 1848794835654279168
author Lehre, Per Kristian
Dang, Duc-Cuong
author_facet Lehre, Per Kristian
Dang, Duc-Cuong
author_sort Lehre, Per Kristian
building Nottingham Research Data Repository
collection Online Access
description The runtime of evolutionary algorithms (EAs) depends critically on their parameter settings, which are often problem-specific. Automated schemes for parameter tuning have been developed to alleviate the high costs of manual parameter tuning. Experimental results indicate that self-adaptation, where parameter settings are encoded in the genomes of individuals, can be effective in continuous optimisation. However, results in discrete optimisation have been less conclusive. Furthermore, a rigorous runtime analysis that explains how self adaptation can lead to asymptotic speedups has been missing. This paper provides the first such analysis for discrete, population-based EAs. We apply level-based analysis to show how a self-adaptive EA is capable of fine-tuning its mutation rate, leading to exponential speedups over EAs using fixed mutation rates.
first_indexed 2025-11-14T19:22:30Z
format Conference or Workshop Item
id nottingham-34365
institution University of Nottingham Malaysia Campus
institution_category Local University
last_indexed 2025-11-14T19:22:30Z
publishDate 2016
recordtype eprints
repository_type Digital Repository
spelling nottingham-343652020-05-04T17:49:55Z https://eprints.nottingham.ac.uk/34365/ Self-adaptation of mutation rates in non-elitist populations Lehre, Per Kristian Dang, Duc-Cuong The runtime of evolutionary algorithms (EAs) depends critically on their parameter settings, which are often problem-specific. Automated schemes for parameter tuning have been developed to alleviate the high costs of manual parameter tuning. Experimental results indicate that self-adaptation, where parameter settings are encoded in the genomes of individuals, can be effective in continuous optimisation. However, results in discrete optimisation have been less conclusive. Furthermore, a rigorous runtime analysis that explains how self adaptation can lead to asymptotic speedups has been missing. This paper provides the first such analysis for discrete, population-based EAs. We apply level-based analysis to show how a self-adaptive EA is capable of fine-tuning its mutation rate, leading to exponential speedups over EAs using fixed mutation rates. 2016-05-27 Conference or Workshop Item PeerReviewed Lehre, Per Kristian and Dang, Duc-Cuong (2016) Self-adaptation of mutation rates in non-elitist populations. In: 14th International Conference on Parallel Problem Solving from Nature, 17-21 Sept 2016, Edinburgh, UK. (In Press)
spellingShingle Lehre, Per Kristian
Dang, Duc-Cuong
Self-adaptation of mutation rates in non-elitist populations
title Self-adaptation of mutation rates in non-elitist populations
title_full Self-adaptation of mutation rates in non-elitist populations
title_fullStr Self-adaptation of mutation rates in non-elitist populations
title_full_unstemmed Self-adaptation of mutation rates in non-elitist populations
title_short Self-adaptation of mutation rates in non-elitist populations
title_sort self-adaptation of mutation rates in non-elitist populations
url https://eprints.nottingham.ac.uk/34365/