Financial Evaluation and Abnormal Data Diagnosis of Listed Enterprises in Public Health Industry

Public health emergencies can have a great impact on the global economy, and at the same time have a major impact on the survival and development of enterprises, and the production and operation of enterprises are hindered. At present, the financial risk of listed companies is frequent, and the financial health of listed companies has become an increasingly prominent problem. To solve the problem of financial health assessment of listed companies, This paper proposes a financial early warning model based on TG‐LSTM network of listed companies based on the TSVD dimensionality reduction method and XGboost ensemble learning model for data optimization, which integrates the LSTM neural network algorithm. This model is proved to effectively assess the financial situation and found abnormal data, through the analysis of the listed company's financial data forecast, after compared with the actual data to verify the validity of the model, through public data sets and other financial data monitoring algorithms of comparative experiments, this model is compared to other models in the Shanghai 50 data set on the MAPE decreased by 0.072%, The MAPE on the CSI 300 dataset is reduced by 0.124 percentage points, and the TG‐LSTM model is proved to have better prediction accuracy. This article is protected by copyright. All rights reserved.