Construction and Simulation of Financial Risk Prediction Model Based on LSTM

• Published in: 2022 6th Asian Conference on Artificial Intelligence Technology (ACAIT) pp. 1 - 6
• Main Author: Nong, Xiaochan
• Format: Conference Proceeding
• Language: English
• Published: IEEE 09.12.2022
• Subjects: Analytical models; Artificial intelligence; data analysis; evaluation indicators; financial risk prediction; LSTM; Predictive models; Simulation
• DOI: 10.1109/ACAIT56212.2022.10137998

To further improve the prediction ability of enterprises in financial risk, a financial risk prediction model based on LSTM is proposed to improve the accuracy of financial risk prediction. The simulation results show that when the dimension and layers of LSTM network are 128 and 4 respectively, the precision of the financial risk prediction model reaches the best level. Compared with Z-score model and Fisher discriminant analysis method, LSTM model shows good performance in F1 measure, precision, accuracy and recall, reaching 93.91%, 94.63%, 94.06% and 95.12% respectively. The above experimental results verify that the financial risk prediction model based on LSTM has feasibility and certain advantages, and has certain practical reference value for the financial risk prediction of enterprises.