Title: Bagging and boosting classification trees to predict churn
Authors: Lemmens, AurĂ©lie
Croux, Christophe
Issue Date: 2003
Publisher: K.U.Leuven - Departement toegepaste economische wetenschappen
Series Title: DTEW Research Report 0361 pages:1-40
Abstract: In this paper, bagging and boosting techniques are proposed as performing tools for churn prediction. These methods consist of sequentially applying a classification algorithm to resampled or reweigthed versions of the data set. We apply these algorithms on a customer database of an anonymous U.S. wireless telecom company. Bagging is easy to put in practice and, as well as boosting, leads to a significant increase of the classification performance when applied to the customer database. Furthermore, we compare bagged and boosted classifiers computed, respectively, from a balanced versus a proportional sample to predict a rare event (here, churn), and propose a simple correction method for classifiers constructed from balanced training samples.
Publication status: published
KU Leuven publication type: IR
Appears in Collections:Research Center for Operations Research and Business Statistics (ORSTAT), Leuven

Files in This Item:
File Status SizeFormat
OR_0361.pdf Published 661KbAdobe PDFView/Open


All items in Lirias are protected by copyright, with all rights reserved.