Churn Detection Using Machine Learning in the Retail Industry

Abstract

The top priority of any business is a constant need Increase sales and profitability. one of the causes of a reduction in profit occurs when an existing customer stops trading. When a customer leaves or terminates the company, potential sales or cross-selling opportunities are lost. When the customer leaves the store without any advice. It can be difficult for companies to respond and take corrective action. Ideally, companies should act proactively and identify themselves chances are you will churn before they leave. customer retention strategies have proven to be less expensive than attracting new ones client. Through data available at the POS(POS) system, extract customer transactions, you can analyze their buying behavior. In this paper Features are created through transactional data and how they are created Identified as important for predicting retail churn industry. Data provided in this document refer to local resident’s supermarket. Thus, dropouts are identified and results are obtained. The results obtained are based on real scenarios. The novelty of this paper is the concept of implementing deep learning algorithms. Convolutional Neural Networks and Restricted Boltzmann Machine is the deep learning technique of choice or restricted Boltzmann machine gave the best results 83% in predicting customer attrition.