Abstract | ||
---|---|---|
In this paper, a novel algorithm for mining maximal frequent patterns is proposed based on projection sum frequent items tree. This algorithm projects the transaction base into a projection sum tree and it can store the frequent itemsets in the tree in a compact manner. The algorithm builds frequent patterns tree directly as FPMax algorithm does. However, all the nodes of PSFIT are sorted and ordered, the children of which are also sorted and ordered. It doesn't need to generate conditional FP-tree dynamically and recursively and it can take advantage of computational result that has been done. The experiment shows that PSFIT is an efficient algorithm, it has comparable performance with FPMax, and in most cases it outperforms FPMax. Key words: projection sum tree, maximal frequent patterns, data mining |
Year | DOI | Venue |
---|---|---|
2007 | 10.1109/FSKD.2007.95 | FSKD (1) |
Keywords | Field | DocType |
data mining | Pattern recognition,Computer science,Algorithm,Artificial intelligence,Database transaction,Machine learning,Recursion | Conference |
Volume | Issue | ISBN |
1 | null | 0-7695-2874-0 |
Citations | PageRank | References |
2 | 0.38 | 7 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Chuanyao Yang | 1 | 11 | 2.32 |
Yuqin Li | 2 | 4 | 1.82 |
Chenghong Zhang | 3 | 116 | 18.03 |
Yunfa Hu | 4 | 74 | 13.44 |