Research Of Stock Analysis And Prediction Based On Grey-Neural Network

With the development of social economy, the stock has become an important investment and financial management tools. Stock Index reflects the overall volatility of the stock on the stock market. Prediction of stock index can help investors to make reasonable decisions. But the stock market is a complex nonlinear system, The traditional methods of statistical analysis and time series prediction is difficult to accurately reflect the characteristics of the stock market. The neural network and support vector machines and other intelligence methods usually require more sample data. For combined models can combine the advantages of the model and overcome the shortcomings of the model, it become a hot research area.Gray neural network model is the combination of GM (1, N) model and BP neural network model. It combines the advantages of neural networks and the gray system model. It can improve forecast accuracy based on a little data. This paper use gray neural network to predict the Shanghai Composite Index, establish the grey neural network model of the daily close index, weekly close index and monthly close index. It was found that with the predicted time span increases, the prediction error is gradually increasing through the prediction research of the daily close index, weekly close index and monthly close index of the Shanghai Composite Index. This is because that as time span increases the volatility of stock index also increased, The gray neural network and the GM (1, N) model is more suitable for stable stock market prediction. Gray neural network model is better than GM (1, N) in prediction accuracy.