Prediction And Application Of Interval Value Of Security Price Based On Recurrent Neural Network And Regression Analysis

With the development of China’s market economy,the securities market plays a more and more important role.The analysis and prediction of the securities price has important theoretical and practical significance,and its research will provide direction for resource allocation.However,due to the influence of many macro and micro factors,it is extremely difficult to realize the accurate prediction of the securities market.Interval value prediction is an important way to predict the stock price.The maximum and minimum values of a period of time give the future fluctuation range of the stock price,so it can better reflect the characteristics of market fluctuation,so as to capture the trading opportunity in the stock market and avoid investment risk.First of all,the volatility of the stock market has the characteristics of linear and nonlinear,the neural network has the ability to approximate the nonlinear function with any precision,and can capture the nonlinear part of the data volatility.However,the extraction ability of the neural network for the linear part is slightly insufficient.Therefore,this dissertation establishes a model combining linear regression and recurrent neural network,which uses the stock correlation theory.Secondly,in the selection of indicators,in addition to the use of basic stock trading data,but also into the technical analysis index data and financial analysis index data,so that the model can cover as much information as possible,then train the neural network model and forecast.Finally,it should be noted that most of the existing research on stock price forecasting use the past information to predict the price of the next day,but these models are unable to show the range of the future rise and fall of the stock price and the best time to buy and sell.This dissertation is different from the previous research results because of it forecasts the highest price and the lowest price in the future,so as to construct the price fluctuation range,and construct the price structure based on the prediction results of the interval value.The results show that the model can improve the prediction accuracy of recurrent neural network,improve the rate of return on investment,and has higher practical reference value for investors.