Predict The Systematic Risk Of The Stock Market Based On Machine Learning Methods

The systemic risk of stock market has been researched and paid close attenti on to for a long time.At present,China’s stock market risk situation is still not o ptimistic,and systemic risk affecting stock market still exists.This paper has done research three aspects to the systematic risk of the stock market.First of all,by standardizing the raw data,calculating the transaction vo lume required during the rising price and fitting the transaction volume and price,we get the relationship between the transaction volume and the price before the systemic risk occurs.Secondly,we merged the standardized data by weeks and months then got a new K-line data K1.And by fitting the K-line data K1 we o btain K-line data K2 which reaches the price peak earlier than the original data.Thirdly,according to the K-Line data K1 and K2,12 features were created.Fur thermore,we use SVR and Linear Regression method to fit the feature data to o btain a linear fitting function that can predict the stock price growth rate in the c oming year.Through real stock data,we find that the three aspects of our work are ef fective,and able to help investors understand the changes in the stock market.Our work also provides a feasible method to the prediction and prevention of systemic risks in the stock market.