WITHDRAWN: REHC: Reduction through exclusive homogeneous clusters for imbalance dataset

Abstract

Abstract Customer retention is a challenging issue of Customer Relationship Management (CRM). To retain a customer it is necessary to predict the customer churn earlier. Retail banking also facing this critical customer churn. Analyzing the customers’ demographic attributes and their behavioral economics enables to predict bank customer churn. The transaction history of customers’ transaction has large data records. Data Reduction on huge volumes of data is certainly required for effective prediction of bank customer churn. This research explores the data reduction techniques that enhances the feature extraction from time series data and further helps for better prediction of customer attrition in banking sector. The proposed approach REHC is helpful to reduce dataset which reflects on time consumption in training the imbalanced dataset and manage CRM effectively by predicting customer churn. It is approached through k-means clustering technique.