Evolutionary algorithm based on Overlapped Gene Expression

Abstract

Inspired by the overlap gene expression in biological study, this paper proposes a novel evolutionary algorithm-EAOGE i.e. Evolutionary Algorithm based on Overlapped Gene Expression. Different from existing works, EAOGE suggests a new expression structure of genes with probabilities of overlapped expression for some segments. The main contributions are: (1) Proposing a novel model and an algorithm of gene expression while borrowing some ideas from artificial immunity algorithm; (2) Analyzing the expressing space and encode characteristic of the new model; (3) The extensive experiments in function finding shows that new model is 2.8∼9.7 times faster than usual GEP method, and in higher-degree polynomial function finding, the success rate of EAOGE is over 10 times than usual GEP. © Springer-Verlag Berlin Heidelberg 2005.