A Frequent Itemset Mining Algorithm based on Interval Intersection Operation
碩士 === 國立臺灣科技大學 === 資訊管理系 === 103 === Frequent Itemset Mining (FIM) is a crucial technique for data mining. This thesis proposes a new FIM algorithm based on interval and bitmap operations. For each itemset in a dataset, an interval set is used to represent the transactions that contain this itemse...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en_US |
| Published: |
2015
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/73802658405251649212 |
| Summary: | 碩士 === 國立臺灣科技大學 === 資訊管理系 === 103 === Frequent Itemset Mining (FIM) is a crucial technique for data mining. This thesis proposes a new FIM algorithm based on interval and bitmap operations. For each itemset in a dataset, an interval set is used to represent the transactions that contain this itemset. Interval intersection operations are then used to find the support counts of the itemset. The experimental results showed that this algorithm takes less execution times than the bit table and Apriori TID algorithms for several configurations with different minimum support thresholds, numbers of transactions, and average lengths of transactions.
|
|---|