Machine Learning Algorithms for Predicting Customers' Lifetime Value: A Systematic Evaluation

With so many successful online merchants in the e-business market in the online era taking heed to customer relationship management (CRM) has become a prerequisite for maintaining competitive advantage. Single customer categories that can be measured in Customer Lifetime Value (CLV) which is defined as the quantity of earnings from a customer in a business during their lifetime, is required to develop an effective CRM strategy. CLV modelling using four most popular Machine Learning(ML) algorithms namely linear regression, decision tree, support vector machine and random forest algorithm are used to find the expected business estimate of a client. The IBM Watson Dataset was used as a training set to evaluate and compare the performance of the four ML regressors in terms of key parameters such as RMSE and R Squared. The findings in this research offer a summary of cutting edge ML methods for CLV prediction.