Research Of Stock Price Prediction Based On Chaotic Time Series And Resilient Feedback Algorithm

With the rapid development of China’s economy, the concept of financial management built upgradually in the public mental, and the stock is the best financial product been favored by people.Stock is one important means of financing of market economy, the development of the stock marketnot only reflects the national economic development, but also related to the vital interests of millionsof households. The nature’s chaos exists everywhere, and the stock market that builds by people mustalso be a chaotic system. Therefore, this article will describe the prediction of stock price movementusing the chaotic time series and feedback neural networks. Details are as follows:First of all, the chaotic dynamics and the theory of chaotic time series are introduced. First, thephenomenon of chaos, definition of chaos, the basic characteristics of chaos, and Lyapunov exponentof chaos-based knowledge are described, followed by a brief description of the chaotic time series,including the phase-space reconstruction technology, time delay and embedding dimension and theprediction using maximum Lyapunov index.Second, the basic of neural networks is introduced. The feedback neural network theory isdescribed in detail, the focus is the feedback neural network algorithms, and the main advantages anddisadvantages are compared. The reason why we choose the prediction method is analyzeddemonstrates ultimately.Finally, the method that chaotic time series and feedback neural network are combined is used topredict the movement of stock price, and the result of this method and the results of prediction usingthe largest Lyapunov exponent and the prediction using classic feedback neural network arecompared.Studies show that the method using the combination of chaotic time series and resilientfeedback algorithm has better effect both in accuracy and performance.