Financial System Crisis Early Warning Model Analysis Based on Neural Network with Entropy Method

Abstract

In order to improve the security of the financial system, an early warning model based on the entropy method was established. Univariate models, multivariate models, quantitative analysis models, and qualitative analysis models were analyzed. The advantages and disadvantages were introduced. It provided the theoretical basis for the following model selection. After that, the indicators and samples of the follow-up model were designed. The accuracy of the new model, the gradual nature of the financial crisis and the new indicators were assumed. The results showed that the neural network model was relatively accurate in both construction and prediction.