Statistical Data Fitting Prediction And Their Applications In Securities Analysis

With rapid economic development of China,development of Stock market tends to Standardization increasingly. From 1991 to now, there has been history in more than ten years. The development of financial market makes Stock exchange record more and more large, which foams a high frequency dataset. Facing high frequency dataset so huge, Can we use it to analyze development trend of Stock effectively? Stock price changes every hour in stock market, How to judge Operation and Stock Price of every company? Especially, for research of the market index will be proposed on the agenda. It is very important to analyze stock market quotation for investor. Meanwhile, there is practical Guiding Significance for the proper operation of stock market.As one of important research topic of data mining, and mining and forecast of time series develop quickly. How to make traditional data mining applied to mining and forecast of time series which is highly concerned by home and abroad Scholars.This paper describes development situation of time series data mining at Home and abroad, and researches forecast and modeling problem of time series. To time series data, what a kind of function should be chosen to be basis function when structure fitting curve? It is decided by requirements of from Complicated dataset mining Unknown and useful information. People first attend to that Polynomial Function could meet demand better in various functions. Polynomial has lots of Advantages when it combined with computer. Polynomial could Evaluation easily, and many more complicated function computation also were approximated by Polynomial. So, it is nothing loss to use Polynomial firstly. In addition, Polynomial has matured in numerical analysis and approximation theory.So, this paper discusses fitting modeling and forecast of basis function. On the basis, SSE composite index is simulated and predicted by us. Prediction results are satisfied, the method is proved that it is effective and simple. We present that some ideal need be improved at last of this paper.