A descriptive encoding language for evolving modular neural networks

Abstract

Evolutionary algorithms are a promising approach for the automated design of artificial neural networks, but they require a compact and efficient genetic encoding scheme to represent repetitive and recurrent modules in networks. Here we introduce a problem-independent approach based on a human-readable descriptive encoding using a high-level language. We show that this approach is useful in designing hierarchical structures and modular neural networks, and can be used to describe the search space as well as the final resultant networks. © Springer-Verlag Berlin Heidelberg 2004.