Research On Application Of Algorithms Of Association Rules

The interesting relations among items of dataset are released by association rule. Current research interesting in the association rule focuses on the algorithm about mining frequent closed itemsets. Yet very little work has been done for updating of frequent closed itemsets. This thesis give an fast updating algorithms of the prevalence CLOSET+.First ,this thesis introduces the concept of Data Mining and Association rule.and then , this thesis presents the conceptions, classes and general thoughts of the algorithms according to the hot research direction of association rule: Frequent itemsets; Frequent closed itemsets; the max frequent itemsets; parallel data mining and distributed mining , and Incremental updating itemsets . An overview on the classical algorithm is given. Also analyse the applicability that the classic algorithm on way of the experiment in the different data.The CLOSET+ is the way of mixture cast shadow strategy, the algorithm is better than other site of the art alorithms on the run time,on memory and expanding way. In this thesis, we introduce an fast updating algorithm of frequent closed itemsets—FUCloset+, which considers the updating of frequent closed itemsets when dynamically adjusting minimum support measure threshold. In worse case, FUCloset+ only scans transaction database once. Moreover, using the previously mined frequent closed itemsets and the item merging, item skipping, sub- itemset pruning methods in CLOSET+, FUCloset+ can obviously improves updating efficiency of frequent closed itemsets. Experimental results show that FUCloset+ algorithm is efficient and effectiveFinally,the example of the data mining in life assurance and traffic accidents is analyze the application of association rule.