Query Learning with Large Margin Classi ersColin

The active selection of instances can sig-niicantly improve the generalisation performance of a learning machine. Large margin classiiers such as support vector machines classify data using the most informative instances (the support vectors). This makes them natural candidates for instance selection strategies. In this paper we propose an algorithm for the training of support vector machines using instance selection. We give a theoretical justiication for the strategy and experimental results on real and ar-tiicial data demonstrating its eeectiveness. The technique is most eecient when the data set can be learnt using few support vectors.