A Comparative Analysis of NFA and Tree-based approach for Infrequent Itemset Mining

Abstract

Frequent itemset mining is an exploratory data mining technique widely used for discovering valuable correlations among data. Frequent itemset mining is a core component of data mining and variations of association analysis, like association rule mining. Extraction of frequent itemsets is a core step in many association analysis techniques. The frequent occurrence of item is expressed in terms of the support count. An item is said to be frequent whose number of occurrences is greater than threshold value. Recently, Infrequent Item sets are also considered to be important in various situations. Infrequent Item set mining is the just the variation of frequent item set mining, where it takes the rarely occurred items in the database. The use of infrequent item set mining is to help for the decision making systems. There are several existing algorithms like Apriori, F-Miner, FP-growth, Residual tree algorithm, Fast algorithm to mine the frequent item sets which takes more computational time. The proposed system is to mine the infrequent item sets by mathematical modeling technique(NFA)where, results in less computing time.