DATA ANALYTICS FOR CUSTOMER RETENTION USING MACHINE LEARNING TECHNIQUES

With rapid increase of competition in business, there is a vital need of analyzing strategies continuously. The client is the focal point of every business. So, it’s compulsory to monitor client’s demands, to make your business profitable and even to retain the client. This study uses logistics regression model to forecast clients loyalty, that whether the client will continue to subscribe the service or will stop. Four imputation techniques were compared to tackle the missing data, which own its own is a big issue. Upon getting the best performing imputation method, two feature reduction techniques were analyzed upon its outcome, to reduce the aspect of variance in the data. Finally, with the high performing imputation method and feature reduction technique, Logistic Regression model was developed, and validated using validation data. After that, predictions were made on test data.