|
Activity Number:
|
507
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics and Marketing
|
| Abstract - #304283 |
|
Title:
|
Incorporating Domain Knowledge in Customer Churn Prediction Using AntMiner+
|
|
Author(s):
|
Wouter Verbeke*+ and Bart Baesens and David Martens and Manu De Backer and Raf Haesen
|
|
Companies:
|
Katholieke Universiteit Leuven and Katholieke Universiteit Leuven and Hogeschool Gent and Hogeschool Gent and Katholieke Universiteit Leuven
|
|
Address:
|
Naamsestraat 69, Leuven, International, B-3000, Belgium
|
|
Keywords:
|
Customer churn prediction ; data mining ; ant colony optimization ; AntMiner+ ; comprehensible rule-sets ; domain knowledge
|
|
Abstract:
|
Customer churn prediction models aim to detect customers with a high propensity to attrite. This paper gives an extended overview of the literature on the use of data mining in customer churn prediction modeling. Furthermore, the novel AntMiner+ classification algorithm is applied to predict churn, and benchmarked. AntMiner+ is a high performing data mining technique based on the principles of Ant Colony Optimization. It allows to include domain knowledge, and seeks to extract intuitive, comprehensible classification rule-sets. Both accuracy and comprehensibility are key aspects of a churn prediction model. Accuracy permits to target future churners in a retention marketing campaign and to improve the efficiency of such campaigns. A comprehensible rule-set on the other hand allows to identify the main drivers for customers to churn and to develop an effective retention strategy.
|
