Computer Science Department
Insurance Risk Evaluation Through Data Mining
Yousaf Nasir Khan
In today’s data driven world, accurate insights from data are of prime importance to the success of insurance companies globally. This project applies data mining techniques to draw insights from a health insurance data set. Firstly, this project explores information available from the data using unsupervised learning. Secondly, significant emphasis is placed on supervised learning to investigate how different data mining algorithms and statistical analysis are useful in predicting the degree of risk associated with a particular applicant for an insurance policy. The algorithms are applied to real data with the goal of producing a predictive model that can accurately classify risk and eligibility of individuals.