MQP 2008- Price and Revenue Optimization


This project applies mathematical techniques to design a process by which an insurance company can optimize revenue. The main objective is to categorize client portfolios using K-means algorithm to segment the market. After clusters are obtained, a logistic regression is applied to determine the optimal premium increase/decrease that maximizes the revenue for each cluster, based its specific characteristics. Applying the optimal premium change to each customer subgroup, the firm will increase its overall revenue.