The Application Of Improved Association Rules And Clustering Algorithm Into Data Mining

With the quickly development of computer, human can get data and store data more quickly, more conveniently, which leads to the amount of data increasing. However, on the basis of survey, only seven percent of these data were exploited. That is to say, people are lost in embarrassed situation of "rich data" and "poor knowledge". Hence, people want to deal with data using computer and get useful and new knowledge.Data mining is the process from which people extract useful and new knowledge. Data mining consists of association rules, clustering analysis, classifying and difference analysis and so on. These algorithms all develop to some extends, but they have some weakness. In the algorithms of association rules of the efficiency of boolean association rules is low and the quality of quantitative association rules is not very good. Moreover, fuzzy clustering algorithms have shortcoming.This paper takes advantage of a character of frequent items set to improve Apriori algorithm, which is a classical algorithm of association rules. This paper gets some improvement in FCM algorithm is that getting cluster center and number of cluster in advance, FCM algorithm clusters data on basis of this result. Last, this paper applies improved FCM algorithm to fuzzy cluster and improved Apriori algorithm to mining association rules.The improved Apriori algorithm has less running time than original Apriori algorithm. And the improved FCM algorithm can avoid getting extreme minimum value. Moreover, the application of improved Apriori algorithm and FCM algorithm to quantitative association rules improves the quality of quantitative association rules.