Advanced Version of A Priori Algorithm
K. Suneetha, Ramanan Krishnamoorti · 2010
Apriori algorithm is a popular classical algorithm of association rule mining. This paper discusses the limitations of the original Apriori algorithm and presents an advanced version of Apriori algorithm to increase the efficiency of generating association rules. Our algorithm uses organized transaction selection approach, where in the rules are generated by picking up the transactions according to the highest order first basis and hence avoiding generation of un-necessary patterns that are not a part of the original database. The major advantage of our approach is that, the number of database scans is massively reduced, to the order of O(n) i.e., of the order of number of transactions available for frequent patterns generation. Hence overcomes the time complexity of original Apriori algorithm which is of the order O(en).