Machine Learning Application in Car Insurance Direct Marketing

Abstract

Direct marketing such as telemarketing or mailing is an important method for companies to boost their business. Identifying the right proportion of target market could largely cut operational expense and improve efficiency.  In this research, a secondary dataset from a car insurance company will be used to study this problem of market targeting. Basing on existing literature study, three classifiers are picked, Naïve Bayesian, Decision Tree, and Neural Network. Some literature researches on each of the algorithm are conducted. Later modelling experiments are performed to predict whether the final customer will purchase the insurance or not.