A Non-sequential Representation of Sequential Data for Churn Prediction

Authors:
Mark Eastwood;Bogdan Gabrys
Affiliations:
Computational Intelligence Research Group, School of Design, Engineering and Computing, Bournemouth University,;Computational Intelligence Research Group, School of Design, Engineering and Computing, Bournemouth University,
Venue:
KES '09 Proceedings of the 13th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems: Part I
Year:
2009

Citing 6
Cited 1

C4.5: programs for machine learning

C4.5: programs for machine learning
Bagging predictors

Machine Learning
Machine Learning for Sequential Data: A Review

Proceedings of the Joint IAPR International Workshop on Structural, Syntactic, and Statistical Pattern Recognition
Pattern Classification (2nd Edition)

Pattern Classification (2nd Edition)
Fast and versatile algorithm for nearest neighbor search based on a lower bound tree

Pattern Recognition
K nearest sequence method and its application to churn prediction

IDEAL'06 Proceedings of the 7th international conference on Intelligent Data Engineering and Automated Learning

Generalised bottom-up pruning: A model level combination of decision trees

Expert Systems with Applications: An International Journal

Quantified Score

Hi-index	0.00

Visualization

Abstract

We investigate the length of event sequence giving best predictions when using a continuous HMM approach to churn prediction from sequential data. Motivated by observations that predictions based on only the few most recent events seem to be the most accurate, a non-sequential dataset is constructed from customer event histories by averaging features of the last few events. A simple K-nearest neighbor algorithm on this dataset is found to give significantly improved performance. It is quite intuitive to think that most people will react only to events in the fairly recent past. Events related to telecommunications occurring months or years ago are unlikely to have a large impact on a customer's future behaviour, and these results bear this out. Methods that deal with sequential data also tend to be much more complex than those dealing with simple non-temporal data, giving an added benefit to expressing the recent information in a non-sequential manner.