An Effective Model for Predicting the Customer Churn in the Retail Sector Based on CRM

Abstract

One of the main problems for large companies is customer churn. Companies are trying to develop methods that can predict potential customers to churn due to the direct impact on retail revenue, especially in the food industry.So, it is significant to identify the factors that enhance customer churn excitement and take the necessary steps to decrease this churn. Hence, in this document, the Levenberg-Marquardt (LM) and ANN (Artificial Neural Network) based prediction model is presented to predict the customer churn in a retail store. Here, the LM algorithm is utilized for the training process of ANN. The outcomesdepict that the proposed prediction model has achieved maximum precision, recall, accuracy and, f-measure as 96.5517241%, 0.918032787 %, 91.1392405% and, 94.1176471 % respectively.Finally,  the results reveals that the presented prediction model has outperformed than other methods for predicting customer churn in a retail store in terms of recall, precision, accuracy, and f-measure.