Abstract:Customer loyalty is one of the key factors in customer retention and segmentation. However, quantitative research in this area based on non-subjective data has been rare in the literature. This paper shows how back propagation neural networks can be used to predict the loyalty level of credit card customers based on transaction records. In the meantime, a set of decision rules are also extracted from the trained neural networks in order to improve the clarity and explicability of the loyalty model.

Predicting Credit Card Customer Loyalty Using Artificial Neural Networks