Improved Algorithm for Frequent Itemsets Mining Based on Apriori and FP-Tree

Abstract

Frequent itemset mining plays an important role in association rule mining The Apriori FP-growth algorithms are the most famous algorithms which have their own shortcomings such as space complexity of the former and time complexity of the latter Many existing algorithms are almost improved based on the two algorithms and one such is APFT 11 which combines the Apriori algorithm 1 and FP-tree structure of FP-growth algorithm 7 The advantage of APFT is that it doesn t generate conditional sub conditional patterns of the tree recursively and the results of the experiment show that it works fasts than Apriori and almost as fast as FP-growth We have proposed to go one step further modify the APFT to include correlated items trim the non correlated itemsets This additional feature optimizes the FP-tree removes loosely associated items from the frequent itemsets We choose to call this method as APFTC method which is APFT with correlation