International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 159 - Number 6 |
Year of Publication: 2017 |
Authors: Balwant Kumar, Dharmender Kumar |
![]() |
Balwant Kumar, Dharmender Kumar . A Matrix based Maximal Frequent Itemset Mining Algorithm without Subset Creation. International Journal of Computer Applications. 159, 6 ( Feb 2017), 23-26. DOI=10.5120/ijca2017912963
Frequent pattern mining is main step in association rule mining. Several algorithms have been proposed for this, but the majority of these algorithms have two main problems that is large number of database scan and generating large candidate itemsets. This process is time intense because these algorithms first mine the minimal frequent itemsets and then generate maximal frequent itemsets from minimal frequent itemsets. Present paper proposes a new top down approach based on compressed matrix for mining maximal frequent itemsets directly without the help of subset. The proposed algorithm performs better than Maximal Frequent Itemset First (MFIF) algorithms with datasets of different size and on different threshold.