Data as voters: instance selection using approval-based multi-winner voting

Abstract: We present a novel approach to the instance selection problem in machine learning (or data mining). Our approach is based on recent results on (proportional) representation in approval-based multi-winner elections. In our model, instances play a double role as voters and candidates. The approval set of each instance in the training set (acting as a voter) is defined from the concept of local set, which already exists in the literature. We then select the election winners by using a representative voting rule, and such winners are the data instances kept in the reduced training set. Our experiments show that, for KNN, the rule Simple 2-EJR (a variant of the Simple EJR voting rule that satisfies 2-EJR) outperforms all the state-of-the-art algorithms and all the baselines that we consider in this paper in terms of accuracy vs reduction. For SVMs, we have obtained slight increases in the average accuracy by using several voting rules that satisfy EJR or PJR compared to the results obtained with the original datasets.