OICC Press International Journal of Mathematical Modelling & Computations 2228-6233 14 1 2024 03 30 Using a New Data Mining Method for Automobile Insurance Fraud Detection: A Case Study by a Real Data from an Iranian Insurance Company 15 20 10.71932/IJM.2024.1126834 EN Maryam Esna-Ashari Insurance Research CenterTehran, Iran Journal Article 2024 03 30 The issue of car insurance fraud is one of the most important issues for insurance companies because it can impose a lot of financial losses on the insurance company. Therefore, timely and early detection of a suspected case can greatly prevent this loss. In the last decade, a lot of studies has been done using data mining techniques in this regard. In this article, we first examine the challenge of imbalanced data, and then, after fixing it, use a very new algorithm introduced in the field of fraud discovery, called XGBoost, for a real data set. Finally, we compare this method with an older method Random Forest algorithm and we will see that the new method works well.