Construction and Application of Mumps Incidence Situation Awareness and Risk Early Warming Model Based on LSTM Algorithm

In order to explore and construct a suitable mumps epidemic situation and risk warning method in Chongqing, 8 characteristic factors were screened out from multi-source data to construct an optimal LSTM model, with an RMSE of 50 and a MAPE of 41. It has been confirmed by the application that the proposed epidemic mumps situation awareness and risk early warning method based on multi-source data has high feasibility, which provides practical reference for the construction of more epidemic situation awareness and risk early warning models of infectious diseases in the future.