Research on The Prediction of Stock Market Based on Chaos and SVM

-The stock market is a very complex system, so it is necessary to use the support vector machine (SVM) algorithm with small sample learning characteristics. The stock market is also a chaotic system, whose financial time series data has chaotic characteristics of random, noise and strong nonlinear. However, the support vector machine for a given time series is usually not considered its chaotic characteristics, so that the regression results are easy to be disturbed and the accuracy is decreased. This paper fully takes into account the chaos characteristics of the stock market, and combines phase space reconstruction theory and support vector machine (SVM). The paper uses C-C algorithm to find the best time delay and the minimum embedding dimension as the input nodes of SVM and establish the chaps-SVM regression model of stock market. The experiment results with SSE (Shanghai Stock Exchange) Composite Index show that, the model can improve the prediction ability of the system with high precision accuracy and good results. This indicates that it is very promising of using the phase space reconstruction theory for economic prediction and it provides an effective method for the prediction of financial time series. Key-Words: -Support vector machine (SVM); Chaos systems; Phase space reconstruction; Prediction