The Research of BP Neural Network Model in Stock Price Forecasting
|School||Qingdao Technological University|
|Course||Applied Computer Technology|
|Keywords||Stock price Neural Networks Stock Prediction Technical Analysis BP algorithm|
With the enhancement of economic development and people 's sense of investment , stock investment has become an important way for many families and personal finance , and to determine a family 's property income . However , the stock market has the characteristics of high-risk and high-yield coexist , and how to create a computing speed and accuracy than the higher stock market forecasting model , with theoretical significance and practical value for investors . Stock prediction method based on the application of neural network technology to predict the stock price , stock operations support topics in-depth exploration and research . Slow learning exist in the stock market forecast for the BP algorithm easy to fall into the local minimum value , the accuracy of predicted results put forward an improved BP neural network algorithm . Take additional momentum term adaptive learning rate measures to speed up the convergence rate of the BP network . BP neural network training sample selection, have a larger impact on the generalization capability of the network , and how to select the appropriate training samples from complex sample data is a difficult . The paper uses technical analysis to select the \The method further improve the generalization capability of the network and the model prediction accuracy . Stock market prediction based on BP neural network principle to establish the stock market prediction model based on BP neural network to take improved BP algorithm for prediction of the stock market , and its forecasting process by MATLAB software simulation experiments . 000,698 of stock prices , for example , to train the prediction model established and trained BP neural network stock prediction model to predict the stock data , to achieve the predicted effect .