The Evolution of Learning Algorithms for Artificial Neural Networks

TL;DR

This paper explores the evolution of neural network learning algorithms using genetic algorithms, demonstrating how local learning rules can be sufficient for learning boolean functions and analyzing emergent learning behaviors.

Contribution

It introduces a genetic algorithm approach to evolve neural networks with local learning rules capable of learning boolean functions, highlighting emergent distributed learning behavior.

Findings

Evolved networks successfully learned boolean functions.

Learning emerges as a distributed property of the network.

Genetic algorithms are effective for discovering neural learning rules.

Abstract

In this paper we investigate a neural network model in which weights between computational nodes are modified according to a local learning rule. To determine whether local learning rules are sufficient for learning, we encode the network architectures and learning dynamics genetically and then apply selection pressure to evolve networks capable of learning the four boolean functions of one variable. The successful networks are analysed and we show how learning behaviour emerges as a distributed property of the entire network. Finally the utility of genetic algorithms as a tool of discovery is discussed.