Information theoretic indicators of fitness, relevant diversity & pairing potential in genetic programming

Stuart W. Card, Chilukuri K. Mohan

Research output: Chapter in Book/Entry/PoemConference contribution

5 Scopus citations

Abstract

Commonly used fitness measures, such as mean squared error, often fail to reward individuals whose presence in the population is necessary to explain substantial portions of the data variance. Diversity indicators are often arbitrary, may reflect diversity irrelevant to solving the problem, and are incommensurate with fitness measures. By contrast, information theoretic functionate are computable general indicators of fitness and diversity without these typical failings. We propose normalized mutual information, redundancy and synergy measures for genetic programming. We also propose selection for recombination and survival by "pairing potential" and "pair potential" estimation, and offer numerical examples as empirical support for theoretical claims.

Original languageEnglish (US)
Title of host publication2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings
Pages2545-2552
Number of pages8
StatePublished - 2005
Event2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005 - Edinburgh, Scotland, United Kingdom
Duration: Sep 2 2005Sep 5 2005

Publication series

Name2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005. Proceedings
Volume3

Other

Other2005 IEEE Congress on Evolutionary Computation, IEEE CEC 2005
Country/TerritoryUnited Kingdom
CityEdinburgh, Scotland
Period9/2/059/5/05

ASJC Scopus subject areas

  • General Engineering

Fingerprint

Dive into the research topics of 'Information theoretic indicators of fitness, relevant diversity & pairing potential in genetic programming'. Together they form a unique fingerprint.

Cite this