A WATER POLLUTION PROBLEM SOLVED: COMPARISON OF GAdC VERSUS OTHER METHODS

摘要

This paper describes a novel method for the supervised training of multivariate regression systems that can be an alternative to other methods such as local regression trees, lazy learners and neural networks. The proposed methodology relies on supervised clustering with genetic algorithms and local learning. Genetic Algorithm driven Clustering (GAdC) offers certain advantages related to robustness, generalization performance, feature selection, explanative behavior, and the additional flexibility of defining the error function and the regularization constraints. Computational results of GAdC are compared with a local regression tree method, lazy learner toolbox and a neural network on an environmental multivariate regression task.