Stock Analysis Based On Data Mining And Prediction

With the market economy of China developing, more and more investors pay attentions to the stock markets. The past one hundred years have witnessed the development and evolution of methodologies regarding analysis and forecast of stocks. However, with the advancement and popularization of computers, these traditional analytical methods have been made public and much commercialized. In addition, the stock markets are increasingly complicated, all these combine to make those methods unilateral. The investors can benefit little from these methods.KDD (Knowledge Discovery in Database) and Data Mining is a new emerging area in the research of artificial intelligence and databases. This technology is used in finance, medical treatment, retail, manufacture, engineering and science. The analysis and forecast of stocks are one of the important areas of KDD. Lots of scholars and corporations devote themselves to the research and application of the KDD in the analysis and forecast of stocks.The thesis discusses how to analyze and forecast the stocks by the basic aspects and the technical aspects of stocks analysis. The methods brought forward are based on the conceptions and technologies in KDD and mathematics. The algorithms and the projects of expriment are expatiated in detail.The main work of the thesis:The financial data are analyzed by using the decision tree classification algorithm ID3. And the author select the typical finicial indexes and make the experiments by using stock samples. Applying these results, the investors can analyze the administrable conditions and the profit capabilities of the corporations.After applying the association rules mining to the discovery of timeseries of stocks, the author add to the constraints of time slices, time intervals and trend rules for the time series of stocks. Analyzing these rules, the investors can better grasp stock movement rules and bargaining timing of stocks.