The research initiative "Data Science and Fraud Detection in Insurance" is a research project bringing together specialists in econometrics, data science and economics, whose objective is to develop new prevention and insurance fraud detection techniques. The project thus contributes to the development and dissemination (in the academic and professional fields) of recent advances in Data science applied to the field of insurance fraud detection.

Automatic fraud detection is a specific area of statistical modeling: unlike the detection of anomalies in the industrial field, for example, it aims to detect fraudulent transactions resulting from the rational behavior of agents. Therefore, detecting cases of fraud requires not only skills in data science and econometrics (knowledge of statistical models and their properties), but also economic and legal skills to understand the motivations and the strategic behavior of fraudsters. The main motivations of the project are mainly related to (i) the modeling of the strategic behavior of fraudsters in the field of insurance; (ii) the need to implement effective detection systems given the huge financial losses associated with fraud; (iii) the use of new databases to identify the mechanisms of insurance fraud. 

For each research action, the work conducted within the chair considers the primary challenges in fraud modeling, encountered both in the academic domain and within Thélem Assurances. In this context, several major issues have been identified: i) the development of new technical tools for the real-time prevention/detection of various types of fraud; 2) the utilization of appropriate methods to handle imbalanced databases; 3) the imperative for evaluating fraud detection models; 4) the demand for interpretability in detection models and the exploration of novel databases.