Show simple item record  

dc.contributor.authorLeathart, Timen_NZ
dc.contributor.authorFrank, Eibeen_NZ
dc.contributor.authorHolmes, Geoffreyen_NZ
dc.contributor.authorPfahringer, Bernharden_NZ
dc.contributor.editorZhang, Min-Lingen_NZ
dc.contributor.editorNoh, Yung-Kyunen_NZ
dc.coverage.spatialSeoul, Koreaen_NZ
dc.date.accessioned2017-11-26T22:28:11Z
dc.date.available2017en_NZ
dc.date.available2017-11-26T22:28:11Z
dc.date.issued2017en_NZ
dc.identifier.citationLeathart, T., Frank, E., Holmes, G., & Pfahringer, B. (2017). Probability calibration trees. In M.-L. Zhang & Y.-K. Noh (Eds.), Proceedings of 9th Asian Conference on Machine Learning (Vol. PMLR 77, pp. 145–160). Seoul, Korea.en
dc.identifier.urihttps://hdl.handle.net/10289/11515
dc.description.abstractObtaining accurate and well calibrated probability estimates from classifiers is useful in many applications, for example, when minimising the expected cost of classifications. Existing methods of calibrating probability estimates are applied globally, ignoring the potential for improvements by applying a more fine-grained model. We propose probability calibration trees, a modification of logistic model trees that identifies regions of the input space in which different probability calibration models are learned to improve performance. We compare probability calibration trees to two widely used calibration methods—isotonic regression and Platt scaling—and show that our method results in lower root mean squared error on average than both methods, for estimates produced by a variety of base learners.
dc.format.mimetypeapplication/pdf
dc.language.isoen
dc.relation.urihttp://proceedings.mlr.press/v77/leathart17a.htmlen_NZ
dc.rights© 2017 T. Leathart, E. Frank, G. Holmes & B. Pfahringer.
dc.sourceACML 2017en_NZ
dc.subjectcomputer scienceen_NZ
dc.subjectProbability calibration
dc.subjectlogistic model trees
dc.subjectlogistic regression
dc.subjectLogitBoost
dc.subjectMachine learning
dc.titleProbability calibration treesen_NZ
dc.typeConference Contribution
dc.relation.isPartOfProceedings of 9th Asian Conference on Machine Learningen_NZ
pubs.begin-page145
pubs.elements-id212095
pubs.end-page160
pubs.finish-date2017-11-17en_NZ
pubs.publication-statusPublisheden_NZ
pubs.publisher-urlhttp://www.acml-conf.org/2017/ACML2017_ConferenceBook.pdfen_NZ
pubs.start-date2017-11-15en_NZ
pubs.volumePMLR 77en_NZ


Files in this item

This item appears in the following Collection(s)

Show simple item record