Skip to main content
WorldCist'18 - 6th World Conference on Information Systems and Technologies

Full Program »

Setting up a Mechanism for Predicting Automobile Customer Defection at SAHAM Insurance (Cameroon)

As markets become more competitive, companies have realized the need to manage the loss of customers (Churn) especially in terms of its prediction. To achieve this, in datamining framework, the main challenge is the selection of variables and the technique adapted to the studied context. This article examines the case of SAHAM insurance and uses ANOVA, chi-square test and Pearson correlations table for variable selection. To make an objective decision on selection of a technique among others, the multi criteria decision aid method PROMETHEE-GAIA has been used. With the aim to improve the initial model, which results was mitigated; the data set has been separated in two groups: individual customers and corporations. Then, with computation of the new one, we observe that, in general, performance is better on the group of individual customers than on previous global model and on corporations.

Rhode Ghislaine Nguewo Ngassam
Catholic University of Central Africa-FSSG-GRIAGES
Cameroon

Jean Robert Kala Kamdjoug
Catholic University of Central Africa-FSSG-GRIAGES
Cameroon

Samuel Fosso Wamba
Toulouse Business School, Département Management de l’Information
France

 

Powered by OpenConf®
Copyright ©2002-2017 Zakon Group LLC