Prediction on customer churn in the telecommunications sector using discretization and Naïve Bayes classifier

In the telecommunications industry, the competitive intensity for retaining existing customers and avoiding losing valuable customers to competitors has increased dramatically. It is a problem of great concern to companies. Customer retention may be boosted by deploying a prediction model to monitor customer activities. In this paper, two experiments with the implementation of data processing techniques using K-Means and Equal-Width Discretization(EWD) combined with Naïve Bayes are performed respectively to conduct a comparison of techniques to identify probable churn activities. Usually, the data generated are of massive size and with highdimensionality. In order to accommodate fast processing, casual heuristics is a preferred deployment. The technique which integrated different algorithm is implemented using Python language under a single processor environment. By using the correlation between attributes, the experimental results show that this can improve the model in identifying the key factors in churn prediction. The results have demonstrated promising overall accuracy.