Gene Expression Programming in Problem Solving

Cândida Ferreira

Departamento de Ciências Agrárias
Universidade dos Açores
9701-851, Angra do Heroísmo, Portugal
candidaf@gene-expression-programming.com
http://www.gene-expression-programming.com/

In this work, the recently invented learning algorithm, gene expression programming, will be introduced focusing mainly on problem solving. Besides a simple introductory example, I chose two relatively complex test problems of symbolic regression. One of these problems was chosen in an attempt to shed some light on the question of constant creation in models discovered with learning algorithms and to provide a palpable measure of the accuracy of the evolved models and the efficiency of the algorithms. The chosen problems also show how gene expression programming is capable of modeling complex realities with great accuracy, allowing, at the same time, the extraction of knowledge from the evolved models.

(In 6th Online World Conference on Soft Computing in Industrial Applications)