Demonstration Analysis For Predicting Stock Price Behavior Owing To Data Mining Technology

This article discussed the application of the data mining technology for analyzing and forecasting Shares. Take data mining technology as direction, this article explore if data mining technology of various kinds can successfully forecasting the reward rate of the A-shares which listed on the Shanghai and Shenzhen stock exchanges. The entire step adopted the process by SEMMA: sample, explore, modify, model, assess, score.The thesis is five parts mainly mark:PartⅠfirstly introduced the relevance concept and the academic personage research about stock market. And then according to classed characteristic of marketplace, it recount the main method of current stock analysis and main part forecasting, and put them under analytical and basically analytical two major kinds of technology. Lastly it narrated Chinese A-share developing history, and summed up the special character in A-share market.PartⅡdescribed the definition of data mining by expert who has narrated home and abroad.Secondly, it carries out classification by the diagram to data mining technology, and then introduced various home and abroad main current makes use of the technology. Finally, it expounded the effect of all kinds date excavating algorithmic by tables in stock analysis and forecast.PartⅢintroduced the fundamental and technology analysis in stock fdrecasting, and then discussed the algorithm about decision-making tree, association analysis, clustering, artificial neural network (ANN), logistic regression, and discussed the application in stock fundamental analysis with the use of them.PartⅣdiscussed the application of decision-making tree, artificial neural network (ANN), time series analysis in stock technology analysis.Part V estimated the forecasting models which are built with every algorithm, and carries out comparison on research results to submit corresponding optimization suggestion.