Fitness landscapes and difficulty in genetic programming

Author

Kinnear, Kenneth E., Jr.

Author_Institution

Adaptive Comput. Technol., Boxboro, MA, USA

fYear

1994

fDate

27-29 Jun 1994

Firstpage

142

Abstract

The structure of the fitness landscape on which genetic programming operates is examined. The landscapes of a range of problems of known difficulty are analyzed in an attempt to determine which landscape measures correlate with the difficulty of the problem. The autocorrelation of the fitness values of random walks, a measure which has been shown to be related to perceived difficulty using other techniques, is only a weak indicator of the difficulty as perceived by genetic programming. All of these problems show unusually low autocorrelation. Comparison of the range of landscape basin depths at the end of adaptive walks on the landscapes shows good correlation with problem difficulty, over the entire range of problems examined

Keywords

algorithm theory; genetic algorithms; learning (artificial intelligence); search problems; adaptive walks; autocorrelation; fitness landscapes; genetic programming; landscape basin depths; landscape measures; random walks; Autocorrelation; Bioinformatics; Computers; Genetic algorithms; Genetic mutations; Genetic programming; Genomics; Visualization;

fLanguage

English

Publisher

ieee

Conference_Titel

Evolutionary Computation, 1994. IEEE World Congress on Computational Intelligence., Proceedings of the First IEEE Conference on

Conference_Location

Orlando, FL

Print_ISBN

0-7803-1899-4

Type

conf

DOI

10.1109/ICEC.1994.350026

Filename

350026