Training method, device and equipment of medical insurance fraud prediction network and storage medium

The method is suitable for the technical field of computers. The invention provides a training method, device and equipment of a medical insurance fraud prediction network, and a storage medium, The method comprises the following steps of: according to the treatment information of the patient sample in the training sample set, acquring a patient feature matrix, analyzing the doctor-patient relationship between the patient and the doctor according to the doctor-seeing information; obtaining a doctor-patient relationship adjacent matrix corresponding to the doctor-patient relationship, accordingto the patient feature matrix and the doctor-patient relationship adjacency matrix, carrying out prediction training on a medical insurance fraud prediction network consisting of a graph convolutional neural network and a variational auto-encoder until a preset training frequency is reached; ending predictive training, Therefore, the characteristics of the associated patient can be well learned through the graph convolut