Development of an Early Warning System for Insurance Fraud Detection Using Statistical Analytics

Abstrak

This study focuses on the development of an Early Warning System (EWS) for detecting insurance fraud using statistical analytics and machine learning approaches. Insurance fraud, particularly false claims, causes substantial financial losses and weakens the credibility of insurance institutions. The objective of this research is to design a proactive detection model capable of identifying fraudulent claims at the earliest possible stage. The study employs a quantitative experimental approach using a labeled dataset of insurance claims. Statistical regression analysis and classification algorithms, including Logistic Regression, Decision Tree, Random Forest, and Support Vector Machine (SVM), were implemented to analyze key variables such as claim amount, claim frequency, and customer behavior. The results show that Random Forest achieved the highest performance, effectively differentiating between fraudulent and legitimate claims. The developed EWS successfully reduced fraud detection time and improved predictive accuracy compared to conventional post-claim investigations. Overall, this research demonstrates that the integration of statistical and machine learning methods provides a more efficient, scalable, and adaptive solution for preventing insurance fraud and safeguarding financial integrity within the insurance sector.

Identifier

• OpenAlex: https://openalex.org/W4416605306
• DOI: https://doi.org/10.70062/jiafc.v1i3.207