Insurance Premium Calculation Using Machine Learning Methodologies
Foad Haji Mohammad · 2023
This research introduces a novel, data-driven optimization methodology for determining insurance premiums and illustrates it using an Iranian health insurance company as a case study.Using the optimal classification machine learning model, the insureds are divided into groups based on their potential to cause disruptions.Next, the indemnity of each group is estimated.Then a linear programming model is used to determine the premium for each group based on their estimated indemnification.Multi-criteria decision-making is utilized to identify the best machine learning algorithm after evaluating the performance of many algorithms using five metrics: accuracy, precision, recall, F1 score, and Matthews correlation coefficient.This study addresses the two common and significant disruptions in the insurance sector: errors in predicting possible losses and a decrease in the insurer's market share.The methodology discussed in this work can be applied to other insurance domains, expanding its practical applications.