Competence-guided editing methods for lazy learning

Abstract

Abstract. Lazy learning algorithms retain their raw training exam-ples and defer all example-processing until problem solving time (eg, case-based learning, instance-based learning, and nearest-neighbour methods). A case-based classifier will typically compare a new tar-get query to every case in its case-base (its raw training data) be-fore deriving a target classification. This can make lazy methods pro-hibitively costly for large training sets. One way to reduce these costs is to filter or edit the original training set, to produce a reduced edited set by removing redundant or noisy examples. In this paper we de-scribe and evaluate a new family of hybrid editing techniques that combine many of the features found in more traditional approaches with new techniques for estimating the usefulness of training exam-ples. We demonstrate that these new techniques enjoy superior per-formance when compared to traditional and state-of-the-art methods. 1