Research on Policy configuration Scheme based on improved ARIMA Model

Abstract

In this paper, we use the principle of different models to remove the trend from the stock time series and choose the SARIMA [1, 1, 1][0,1,0] 12 models to fit the stock price. The results show a good prediction performance. Then a ternary transaction strategy is proposed, accurately identifying value changes regarding cash as financial products (expected rate of return = 0%, cost = 0%), and establishes a ternary asset flow model. Combining the expected rate of return and volatility, we model the value of existing financial products. Based on the trend accuracy to evaluate the accuracy of stock price prediction, the SARIMA model shows high accuracy in the verification set, and the maximum root mean square error (RMSE) is 11. 5494. The final five-year return is 51973 times higher than the bank savings and fixed investment return.