Mining Approximate Frequent Itemsets Using Pattern Growth Approach

Approximate frequent itemsets (AFI) mining from noisy databases are computationally more expensive than traditional itemset mining. This is because the AFI algorithms generate large number of candidate itemsets. article proposes an algorithm to mine AFIs using pattern growth approach. The major contribution proposed approach it mines core patterns and examines approximate conditions directly with single phase two full scans database. Related apply Apriori-based generation test require multiple phases obtain complete AFIs. First generates patterns, second patterns. Specifically, novel techniques that how map transactions on FP-tree, conditional FP-tree. FP-tree maps shared branches when share a similar set items. reduces size helps efficiently compute We compare performance our state art benchmark databases. experiments analyzed by comparing processing time scalability varying database transaction length. results show in less related algorithms.