Analysis Automobile Insurance Fraud Claim Using Decision Tree and Random Forest Method

https://doi.org/10.47194/ijgor.v5i4.337

Authors

Abstract

Insurance fraud, particularly in the automobile sector, poses significant financial risks to insurance companies. This study aims to analyze fraudulent claims in automobile insurance using Decision Tree and Random Forest methods. A dataset consisting of 10,000 entries was utilized, containing variables such as vehicle type, claim amount, and claim status. The Decision Tree method was employed for its interpretability, while Random Forest was used for its superior accuracy. Results indicated that the Random Forest model outperformed the Decision Tree model, achieving an accuracy of 51.37% compared to 50.47%. This research highlights the effectiveness of machine learning techniques in detecting insurance fraud and provides insights for insurers to enhance their fraud detection systems.

Published

2024-12-16