Securities Market Analysis And Research Of The Decision-making Model

The stock market is a place to provide financing function, any institution, companies or individuals can finance by transaction of stocks or other securities. In order to maintain and even strengthen the financing function of the security market, the SFC and other regulatory authorities should control the overbought and oversold of stock market to make the stock market development which could keep the majority of institutions or individual investor keep the interested in stock market. Thus, analyzing the relation between the inherent law and the factors impaction of stock to make the development of our country's stock market even more standard so as to keep the financing function is the problem that the paper to solve.Take the stock in stock market as the example, the paper analyze the stock market and establish the control model, then describe the relation between impaction factors and stock market. The paper will analyze the stock by knowledge of automation and control which is the foundation of proposed BP neural network analysis. Compare to neural network analysis and other analyze method, the formal analyze method has a good application effect for its nonlinear approximation ability, self-learning and self-adaption. This is because BP network can automatically identify the inherent laws of stock market movement on the basis of historical data, and use these laws to adjust weights and valve values of the neural network. This feature is good in promoting of control the development of stock market to keep the financing function. The first key point in this paper is that we propose an improved BP algorithm owing to some limitations of BP network in application. This improved BP algorithm can increase the convergence rate by increasing the error offset of back-propagation, and this method has been proved that the effect is obviously good.This paper analyze the relationship between CPI, M1, number of shares outstanding, savings deposits and stock market among the factors that affect the stock market. This paper selects six groups of data contains above four factors as input variables of a decision-making model. Then we select the various relevant parameters of BP neural network and the objective function through experiments. The second key point in this article is that we establish a new model of stock market control after much more tests, and then obtain the impact situation of each variable on the stock market by the debugging of input variables.The real effect of this model needs to be validated in practice, but the affection of variety of economic variables from the analysis by the model to stock market will take a promote effect to the healthy development of stock market and financing function.