A Spatio-temporal Neural Network for Medical Insurance Fraud Detection

Medical insurance fraud presents a significant challenge to the medical insurance ecosystem and is one of the major issues that urgently need to be addressed in the current field of medical insurance. In recent years, due to the growing complexity of fraudulent methods in medical insurance, existing fraud detection methods are unable to effectively tackle this challenge. To this end, we innovatively introduce spatio-temporal medical data into the model and propose a Spatio-Temporal Neural Network (STNN) method to identify fraud in medical insurance. This technique analyzes health records by combining spatial and temporal dynamics through three-dimensional convolution. We evaluate STNN on an authentic dataset of medical insurance. STNN outperforms state-of-the-art benchmarks in terms of precision-recall evaluations, highlighting its efficacy as a powerful tool in combating fraudulent activities within the medical insurance domain.