Summary: Fitness Distance Correlation Analysis:
An Instructive Counterexample
Lee Altenberg
Hawaii Institute of Geophysics and Planetology
University of Hawaii at Manoa, Honolulu, HI USA 96822
WWW: http://pueo.mhpcc.edu/ ” altenber/
c
fl1997 by Lee Altenberg. Expanded from the version in Baeck, T. ed., Proceedings of the
Seventh International Conference on Genetic Algorithms (ICGA97), Morgan Kaufmann, San Francisco, CA.
Fitness distance correlation (FDC) has been of­
fered as a summary statistic with apparent suc­
cess in predicting the performance of genetic al­
gorithms for global optimization. Here, a coun­
terexample to Hamming­distance based FDC is
examined for what it reveals about how GAs
work. The counterexample is a fitness func­
tion that is `GA­easy' for global optimization,
but which shows no relationship between fitness
and Hamming distance from the global optimum.
Fitness is a function that declines with the number

Source: Altenberg, Lee - Department of Information and Computer Science, University of Hawai'i at Manoa

Collections: Computer Technologies and Information Sciences