Stock Price Forecast And Analysis Based On LASSO Method And Neural Network

With the rapid development of the economy,the stock market is constantly improving.Nowadays,stocks have become one of the most important investment methods for investors.However,stocks are highly profitable and risky.How to effectively analyze stocks and analyze stocks Price prediction is of great significance to investors.This article uses the LASSO method to construct a stock price prediction model,and uses the historical data of Changchun High-tech Stock(stock code 000661)to conduct an empirical study.A reasonable analysis of the stock market,and according to its own characteristics,puts forward a stock price analysis method combining BP neural network(Back Propagation)and LASSO method.By selecting the historical data of a stock price,and then performing regression screening through the LASSO method,the input variables are optimized,the number of input neurons is reduced,the calculation complexity is reduced,and the input variables and single The model's network training data is compared.The experimental results are obtained.The average absolute error between the true value and the predicted value of the four experiments is 4.19%,6.68%,6.53%,and 5.55% respectively.According to the experimental results,it can be known that the prediction accuracy based on the combination of LASSO method and BP is significantly higher than that of BP neural network,and the correlation coefficient and linear regression are respectively different from the model accuracy of the combination with BP.Therefore,the model based on the combination of LASSO method and BP neural network can realize effective prediction of stock prices,and has certain reference value for stock investment.