Ensembles of Probability Estimation Trees for Customer Churn Prediction

Customer churn prediction is one of the most, important elements tents of a company's Customer Relationship Management, (CRM) strategy In tins study, two strategies are investigated to increase the lift. performance of ensemble classification models, i.e (1) using probability estimation trees (PETs) instead of standard decision trees as base classifiers; and (n) implementing alternative fusion rules based on lift weights lot the combination of ensemble member's outputs Experiments ale conducted lot font popular ensemble strategics on five real-life chin n data sets In general, the results demonstrate how lift performance can be substantially improved by using alternative base classifiers and fusion tides However: the effect vanes lot the (Idol cut ensemble strategies lit particular, the results indicate an increase of lift performance of (1) Bagging by implementing C4 4 base classifiets. (n) the Random Subspace Method (RSM) by using lift-weighted fusion rules, and (in) AdaBoost, by implementing both