A Fuzzy Markov Model for Consumer Credit Behavior Dynamics

A heterogeneous fuzzy Markov model is built for modeling and forecasting credit behavior dynamics of credit card users in this paper. This model brings together a fuzzy rule-based inference system and the Markov chain. The credit migration rate matrix of the fuzzy Markov chain is obtained by the inference system based on reasonable setting of rules for each consumer, and updated at each time interval, a process which results in heterogeneity across consumers and over time, thereby imparting more flexibility and accuracy to the model to reflect the reality. Credit status evolution process is predicted for a simulated dataset of consumer credit behaviors.