Churn Prediction in Banking System using K-Means, LOF, and CBLOF

Customer churn prediction helps in identifying those customers who are probable to stop a subscriptions, products or services, and is therefore very essential for any business. Churn predictions can be very valuable for customers retentions, as it helps in predicting customer that are at risks of sendoff. It is more challenging to put forth churn prediction in banking sector, as there are no contractual agreements between a customer and a bank regarding the duration of services. Loss of customers can be very costly as it is very expensive to obtain new customers in this age of competition. There are many churn prediction techniques however; K-Means, Local Outlier Factors (LOF) and Cluster-Based Local Outlier Factors (CBLOF) have not been used so far for this purpose. In this research, we apply these techniques for customer churn prediction. The results are evaluated and analyzed using Precision (Pr), Recall (Re) and F1-Measures to justify the efficiency and effectiveness of this research.