Application and Innovation in Financial Market Risk Management Based on Big Data and Artificial Intelligence

Abstract Finance is the core of the economy, and the size of its risk is directly related to the immediate interests of the people. This paper combines the traditional measurement GARCH family model, the common time series model-ARFIMA model and the LSTM deep neural network in tandem, and constructs the ARFIMA-GARCH-LSTM hybrid model to predict the financial market risk management data. Then the VAR model is used to measure the risk value of the financial market to realize the accurate management of the financial market risk. The prediction error of this paper’s model is concentrated within 0.002-0.005, which is small, and the value of the prediction index under the tandem combination method is less than 0.02. It shows that this paper’s prediction model has a good effect on predicting financial market risks. When the value at risk is below the warning line, the VaR risk assessment model detects that the real rate of return is very low continuously, which achieves better results, based on which this paper proposes corresponding preventive measures of financial market risk, which makes a certain contribution to the security of the economy and society.