Financial risk prediction method, device and storage medium

Abstract

The invention discloses a financial risk prediction method, comprising: obtaining financial data of sampled customers; calculating the efficacy coefficient value of the risk evaluation index of the sampled customer according to the financial data, and calculating the risk weight of each risk evaluation index through the entropy weight method; according to the risk The efficacy coefficient value and risk weight of the evaluation index calculate the financial risk score of the sampled customers, and define the risk level; perform neural network training according to the sliding cycle, the efficacy coefficient value of the risk evaluation index, risk weight, and risk level, and generate a neural network prediction model; The current financial risk index of the target customer is input into the neural network prediction model to obtain the financial risk level of the target customer. The financial risk prediction method disclosed by the invention can carry out weight division for different risk evaluation indexes and improve the accuracy of financial risk prediction. The invention also discloses a financial risk forecasting device and a storage medium.

Description

technical field [0001] The present invention relates to the technical field of financial risk forecasting, in particular to a financial risk forecasting method, device and storage medium. Background technique [0002] At present, most of the objects of risk early warning models are listed companies, and most of them use multivariate models. [0003] In the prior art, multivariate models include logistic regression models, multiple linear decision models, Z-score models, and the like. [0004] The inventor finds in the process of implementing the present invention that there are the following technical problems in the prior art: [0005] The above models all use bankruptcy or financial losses for two consecutive years as the most serious financial risk. The risk judgment indicators are relatively simple, and the influence of financial data in the company's development process is not considered, resulting in a low accuracy rate of the prediction model. Contents of the inven...

AI Technical Summary

Problems solved by technology

[0005] The above models all use bankruptcy or financial losses for two consecutive years as the most serious financial risk. The risk judgment indicators are relatively single, and the impact of financial data in the company’s development process is not considered, resulting in a low accuracy rate of the prediction model.

Images