Towards a runtime comparison of natural and artificial evolution Journal Article

Author(s): Paixão, Tiago; Pérez Heredia, Jorge; Sudholt, Dirk; Trubenová, Barbora
Article Title: Towards a runtime comparison of natural and artificial evolution
Affiliation IST Austria
Abstract: Evolutionary algorithms (EAs) form a popular optimisation paradigm inspired by natural evolution. In recent years the field of evolutionary computation has developed a rigorous analytical theory to analyse the runtimes of EAs on many illustrative problems. Here we apply this theory to a simple model of natural evolution. In the Strong Selection Weak Mutation (SSWM) evolutionary regime the time between occurrences of new mutations is much longer than the time it takes for a mutated genotype to take over the population. In this situation, the population only contains copies of one genotype and evolution can be modelled as a stochastic process evolving one genotype by means of mutation and selection between the resident and the mutated genotype. The probability of accepting the mutated genotype then depends on the change in fitness. We study this process, SSWM, from an algorithmic perspective, quantifying its expected optimisation time for various parameters and investigating differences to a similar evolutionary algorithm, the well-known (1+1) EA. We show that SSWM can have a moderate advantage over the (1+1) EA at crossing fitness valleys and study an example where SSWM outperforms the (1+1) EA by taking advantage of information on the fitness gradient.
Keywords: theory; Population Genetics; Natural evolution; Runtime analysis; Strong selection weak mutation regime; Evolutionary algorithms
Journal Title: Algorithmica
Volume: 78
Issue 2
ISSN: 1432-0541
Publisher: Springer  
Date Published: 2017-06-01
Start Page: 681
End Page: 713
Copyright Statement: CC BY 4.0
Sponsor: The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under Grant Agreement No. 618091 (SAGE).
DOI: 10.1007/s00453-016-0212-1
Open access: yes (OA journal)
IST Austria Authors
  1. Tiago Paixão
    27 Paixão