Using Adaptive Operators in Genetic Search

Jonatan Gómez¹, Dipankar Dasgupta², and FabioGonzález¹

¹Division of Computer Science
The University of Memphis
Memphis TN 38152
and Universidad Nacional de Colombia
Bogotá, Colombia
{jgomez,fgonzalz}@memphis.edu

²Division of Computer Science
The University of Memphis
Memphis TN 38152
dasgupta@memphis.edu

Abstract. In this paper, we provided an extension of our previous work on adaptive genetic algorithm [1]. Each individual encodes the probability (rate) of its genetic operators. In every generation, each individual is modified by only one operator. This operator is selected according to its encoded rates. The rates are updated according to the performance achieved by the offspring (compared to its parents) and a random learning rate. The proposed approach is augmented with a simple transposition operator and tested on a number of benchmark functions.

LNCS 2724, p. 1580 ff.

Full article in PDF

lncs@springer.de
© Springer-Verlag Berlin Heidelberg 2003