Program      Committee      Author Index      Search      About GECCO      CD Tech Support

Session: Workshop - Learning Classifier Systems (LCS) Title: Empirical evaluation of ensemble techniques for a Pittsburgh Learning Classifier System     Authors: Jaume Bacardit Natalio Krasnogor     Abstract: Ensemble techniques have proved to be very useful to boost the performance of several types of machine learning methods. In this paper, we illustrate its usefulness in combination with GAssist, a Pittsburgh-style Learning Classifier System. Two types of ensemble are tested. First bagging-style consensus prediction. Second an ensemble intended to deal more efficiently with ordinal classification problems. Both methods improve the performance and behaviour of GAssist in the tested domains.