BoostML: An adaptive metric learning for nearest neighbor classification

Simple item page
dc.contributor.authorZaidi, N.
dc.contributor.authorSquire, D.
dc.contributor.authorSuter, D.
dc.contributor.conferencePacific-Asia Conference on Knowledge Discovery and Data Mining (14th : 2010 : Hyderabad, India)
dc.contributor.editorZaki, M.J.
dc.contributor.editorYu, J.X.
dc.contributor.editorRavindran, B.
dc.contributor.editorPudi, V.
dc.date.issued2010
dc.description.abstractA Nearest Neighbor (NN) classifier assumes class conditional probabilities to be locally smooth. This assumption is often invalid in high dimensions and significant bias can be introduced when using the nearest neighbor rule. This effect can be mitigated to some extent by using a locally adaptive metric. In this work we propose an adaptive metric learning algorithm that learns an optimal metric at the query point. We learn a distance metric using a feature relevance measure inspired by boosting. The modified metric results in a smooth neighborhood that leads to better classification results. We tested our technique on major UCI machine learning databases and compared the results to state of the art techniques. Our method resulted in significant improvements in the performance of the K-NN classifier and also performed better than other techniques on major databases.
dc.description.statementofresponsibilityNayyar Abbas Zaidi, David McG. Squire and David Suter
dc.identifier.citationProceedings of the 14th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD 2010), held in Hyderabad, India, 21-24 June 2010: pp.142-149
dc.identifier.doi10.1007/978-3-642-13657-3_17
dc.identifier.isbn9783642136573
dc.identifier.issn0302-9743
dc.identifier.issn1611-3349
dc.identifier.orcidSuter, D. [0000-0001-6306-3023]
dc.identifier.urihttp://hdl.handle.net/2440/63428
dc.language.isoen
dc.publisherSpringer-Verlag
dc.publisher.placeGermany
dc.relation.ispartofseriesLecture notes in Computer Science ; 6118
dc.rightsCopyright Springer-Verlag Berlin Heidelberg 2010
dc.source.urihttp://www.springerlink.com/content/978-3-642-13671-9/?k=suter
dc.subjectAdaptive Metric Learning
dc.subjectNearest Neighbor
dc.subjectBias-Variance analysis
dc.subjectCurse-of-Dimensionality
dc.subjectFeature Relevance Index
dc.titleBoostML: An adaptive metric learning for nearest neighbor classification
dc.typeConference paper
pubs.publication-statusPublished

Files

Collections

Computer Science publications