The Study Of Stock Price Index Prediction Based On Grey Theory And Neural Network Theory

Stock market is a complex non-linear dynamic system. It is difficult to reflect market with the trait of more factors,non-linear and time variety using the traditional timing prediction technology. Based on the review of traditional prediction methods,the paper forward a system composed of grey theory and neural network theory,and an ameliorative method on its function is studied.Introduction depict this paper's main content and significance of research. Then, this paper introduced the securities forecasting theory,summaries stock forcasting method and present condition and existent problem.Then,this paper introduced the basic knowledge of the grey correlation analysis and grey models,compared various models in forecasting stock price index. This paper use Shanhai Stock Exchange Comprehensive index, because the index reflect the overall price level of the market to the greatest extent , whose signal most strongly influenced the investors. The index can reflect the domestic stock market trends exactly, having high predictive value and better predictability. Empirical results indicate that the Grey-Markov chain model has high accuracy and the value of application. This is the main innovation of the paper.Finally,this paper introduced Nerual Network theory and learning arithmetic of BP Network. The paper forward Nerual Network forecasting system based on grey correlation analysis. The way breaks a new path to further definition of dynamic stock market. The paper put forward to use the thought of grey correlation into network train process to regulate the number of the implicit node to realizes the ability of superior of the network to attain the better prediction result. Grey relationship analysis theory is used to filter the most important quantitative technical indices which can reflect stock price tendency in order to optimize the input parameters of the BP Neural Network. The conclusion shows the new system of stock index prediction can provide good prediction for this problem. And set up an effective analytical system for vast investors. This chaper is important and innovation.At last,the paper summaries the fruits of this paper and gives the prospect of future work.