Self-adaptation of Mutation Rates in Non-elitist Populations

被引:52
作者
Dang, Duc-Cuong [1 ]
Lehre, Per Kristian [1 ]
机构
[1] Univ Nottingham, Sch Comp Sci, Nottingham, England
来源
PARALLEL PROBLEM SOLVING FROM NATURE - PPSN XIV | 2016年 / 9921卷
关键词
ALGORITHMS;
D O I
10.1007/978-3-319-45823-6_75
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
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 levelbased 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.
引用
收藏
页码:803 / 813
页数:11
相关论文
共 14 条
[1]  
Back T., 1992, Proceedings of the 1st European Conference on Artificial Life, P263
[2]  
Böttcher S, 2010, LECT NOTES COMPUT SC, V6238, P1, DOI 10.1007/978-3-642-15844-5_1
[3]  
Corus D, 2014, LECT NOTES COMPUT SC, V8672, P912
[4]   Refined Upper Bounds on the Expected Runtime of Non-elitist Populations from Fitness-Levels [J].
Dang, Duc-Cuong ;
Lehre, Per Kristian .
GECCO'14: PROCEEDINGS OF THE 2014 GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE, 2014, :1367-1374
[5]   Optimal Parameter Choices Through Self-Adjustment: Applying the 1/5-th Rule in Discrete Settings [J].
Doerr, Benjamin ;
Doerr, Carola .
GECCO'15: PROCEEDINGS OF THE 2015 GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE, 2015, :1335-1342
[6]   Solving Problems with Unknown Solution Length at (Almost) No Extra Cost [J].
Doerr, Benjamin ;
Doerr, Carola ;
Koetzing, Timo .
GECCO'15: PROCEEDINGS OF THE 2015 GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE, 2015, :831-838
[7]  
Eiben AE, 2007, STUD COMPUT INTELL, V54, P19
[8]   Complete genetic linkage can subvert natural selection [J].
Gerrish, Philip J. ;
Colato, Alexandre ;
Perelson, Alan S. ;
Sniegowski, Paul D. .
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA, 2007, 104 (15) :6266-6271
[9]   Completely derandomized self-adaptation in evolution strategies [J].
Hansen, N ;
Ostermeier, A .
EVOLUTIONARY COMPUTATION, 2001, 9 (02) :159-195
[10]  
Lehre P. K., 2013, Foundations of Genetic Algorithms, FOGA 2013,, P97