Prediction Of The Share Price Based On Wavelet Analysis And Artificial Neural Network

In the stock market , stock prices are the focus of the shareholders. It is necessary to perform scientific and mathematical analysis, identify the trends and the key factors, guiding investors to the stock investments. It would be not so suitable and appropriate to apply general timing analysis methodology in describing the alteration of stock prices thanks to its macro, micro and uncertainty. In recent years, two hot mathematical tools—wavelet analysis and artificial neural networks were founded, which can effectively describe the trend of stock prices quotation, and are proved to be effective in the analysis of the price curve.In this paper, wavelet analysis and artificial neural networks are used as two mathematical tools to analyze in stock prices and forecasts. It introduces the objective, significance and research methods of this study . Then it presents the role that the wavelet analysis and artificial neural network theory play in the forecasts of stock price. As regards the problems of neural network ,the author analyzes and has tried to solve it. Wavelet analysis and Artificial neural networks are used for analysis about the price of selected stocks .After discounting the noise of the price curve, it becomes generally smooth,and then used to forecast and make training. Finally we apply instances in studying Chinese stock market with a good result. Compared to other means such as ARIMA model, we conclude that WA-ANN method is a good one in the stock prediction.