Customizing Insurance Product Based On Customer Data Leveraging Learning Algorithms

With the increasing number of customers willing to trade their data in exchange for lower premiums, harnessing customer and insurance data to generate a customized insurance product is a powerful concept. Personalization engines have very often been powered by machine learning algorithms and big data analytics. Customer data once collected, can be analyzed for insights and then utilized in decision making with the help of off-the-shelf or novel learning algorithms. One common barrier in this process is the problem of imbalanced insurance data which affects the accuracy of these algorithms. We have carried out a thorough analysis of the applications of various learning algorithms such as Random Forest, Bayesian Network, Neural Network, Apriori algorithm and so on. Through our paper, we aim to present a clear picture of the advancements in systems that deliver customized insurance products to customers while discussing how all of those approaches can be used to solve problems pervasive in the industry.