Please use this identifier to cite or link to this item:
Full metadata record
DC FieldValueLanguage
dc.contributor.authorPesantez-Narvaez, Jessica-
dc.contributor.authorGuillén, Montserrat-
dc.contributor.authorAlcañiz, Manuela-
dc.description.abstractA boosting-based machine learning algorithm is presented to model a binary response with large imbalance, i.e., a rare event. The new method (i) reduces the prediction error of the rare class, and (ii) approximates an econometric model that allows interpretability. RiskLogitboost regression includes a weighting mechanism that oversamples or undersamples observations according to their misclassification likelihood and a generalized least squares bias correction strategy to reduce the prediction error. An illustration using a real French third-party liability motor insurance data set is presented. The results show that RiskLogitboost regression improves the rate of detection of rare events compared to some boosting-based and tree-based algorithms and some existing methods designed to treat imbalanced responses.-
dc.format.extent21 p.-
dc.relation.isformatofReproducció del document publicat a:
dc.relation.ispartofMathematics, 2021, vol. 9, num. 579, p. 1-21-
dc.rightscc-by (c) Pesantez-Narvaez, Jessica et al., 2021-
dc.sourceArticles publicats en revistes (Econometria, Estadística i Economia Aplicada)-
dc.subject.classificationAnàlisi de regressió-
dc.subject.classificationTeoria de l'estimació-
dc.subject.classificationProcessament de dades-
dc.subject.classificationSistema binari (Matemàtica)-
dc.subject.otherRegression analysis-
dc.subject.otherEstimation theory-
dc.subject.otherData processing-
dc.subject.otherBinary system (Mathematics)-
dc.titleRiskLogitboost Regression for Rare Events in Binary Response: An Econometric Approach-
Appears in Collections:Articles publicats en revistes (Econometria, Estadística i Economia Aplicada)

Files in This Item:
File Description SizeFormat 
710731.pdf4.43 MBAdobe PDFView/Open

This item is licensed under a Creative Commons License Creative Commons