Evolving neural networks for chlorophyll-a prediction

Abstract

The paper studies the application of evolutionary artificial neural networks to chlorophyll-a prediction in Lake Kasumigaura (in Japan). Unlike previous applications of artificial neural networks in this field, the architecture of the artificial neural network is evolved automatically rather than designed manually. The evolutionary system is able to find a near optimal architecture of the artificial neural network for the prediction task. Our experimental results have shown that evolved artificial neural networks are very compact and generalise well. The evolutionary system is able to explore a large space of possible artificial neural networks and discover novel artificial neural networks for solving a problem.