Study On The Application Of Association Rules In Temporal Data Mining

In the later 1980's, Data Mining-a new research field, appears gradually and develops rapidly. The study purpose of Data Mining is to find the regular information which hided in large data set and people interested in. As one of the mining objects of Data Mining, temporal data base is composed by series of sequence or affair. The study of temporal data mining plays an important role in the data analysis of Commerce, Finance, Science and Engineer. So the methods of temporal data mining became a research hotspot of Data Mining.Association rules is always a hot research subject in Data Mining and Artificial Intelligence which has already been applied in the areas of CRM, Physic, Biology and so on. Traditional Association rules mining processes like Market Basket Analysis seldom consider the temporal constrain. Since the scale and importance of temporal data base ceaselessly grandness, how to apply Association rules mining into temporal data base became a subject which highly worth our research. Besides that, the application of Association rules in security market is still at the start stage. How to effectively use Association rules mining in the data analysis of security tradesystem also need our study.Start form the background of the study this paper expatiate the technology of Data mining and status quo of temporal data mining. It also introduce the theory of association rules and its application in temporal data mining.On the basis of deeply analyze the temporal association rules mining, the paper improves the mining arithmetic of commonly temporal association rules, brings forward Weighted Temporal Association Rules and study its feasibility of application. Using the improved method to mine the temporal data base, the results can finely detect the diversification of purchasing custom of customers and offer supports to the marketing decision making. The paper also compares and analyzes several results by different association rules mining models and verifies the validity and rationality of weighted theory.The paper also innovatively brings forward a method of Temporal Constrained Quantitative Association Rules Mining and uses it into the validity verifying of the technical analysis indexes in stock market. The paper chooses the Relative Strength Index (RSI), collects trade data to do demonstration analysis and mines out several useful quantitative association rules which can guide technical analysis application and actual operation in investment.