Predictive Maintenance In Predicting Machine Failures At Unilever: A Log-Based Approach Using Machine Learning Algorithm

Abstract

Identifying customer churn plays a vital role in the operation and expansion of any business. By identifying non-returning consumers, a company can learn why customers leave and can then tailor their marketing campaigns to maximize business growth. business. This study aims to detect and implement a machine learning model that can almost accurately analyze and have customer churn prediction in the E-commerce Retail industry. With the increasing recognition and rapid development of the e-commerce industry, e commerce retail began to compete more fiercely (Zhang, 2015). Today, the most important option is how to retain consumers, which can be implemented by providing good services and benefits at a reasonable price. Therefore, with the growing e-commerce platform in Vietnam and increasing competition among companies when meeting the needs of customers, having a strategy and implementing a variety of service can attract customers and keep them stay. In addition, the non-return of customers on the e-commerce platform represents unbalanced consumer categories and difficult-to-identify changes. This imbalance is due to the fact that the number of customers leaving is much larger than the number of customers not leaving. This study analyzes the causes and proposes methods by building a supervised machine learning model to identify the main factors in customer abandonment, thereby helping businesses to predict ahead of time customer behaviors, quickly have plans to improve and retain customers.