Coevolutionary Learning: A Case Study

Abstract

Coevolutionary learning, which involves the embedding of adaptive learning agents in a fitness environment that dynamically responds to their progress, is a potential solution for many technological chicken and egg problems. However, several impediments have to be overcome in order for coevolutionary learning to achieve continuous progress in the long term. This paper presents some of those problems and proposes a framework to address them. This presentation is illustrated with a case study: the evolution of CA rules. Our application of coevolutionary learning resulted in a very significant improvement for that problem compared to the best known results. 1 Introduction A recurrent issue in the field of machine learning is that the performance of a learning system relies heavily on the amount of knowledge that has been introduced by the designer. This knowledge can be expressed in the form of an appropriate representation, specific search operators, a training set which provides a good...