Modifying iEclat algorithm for infrequent patterns mining
Pattern ruining has been extensively studied in research due to its successful application in several data mining scenarios. Association rules mining is a basic step to determine the correlation between data items based on frequency of occurrence. In database, data items can be found as frequent p...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English |
| Published: |
2018
|
| Subjects: | |
| Online Access: | http://eprints.unisza.edu.my/1157/1/FH03-FIK-20-39606.pdf http://eprints.unisza.edu.my/1157/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Pattern ruining has been extensively studied in research due to its successful application in several data mining
scenarios. Association rules mining is a basic step to determine the correlation between data items based on frequency
of occurrence. In database, data items can be found as frequent pattern and infrequent pattern. Frequently occuring
pattern has been an interesting issue of research in marketing for the past 24 years. However, infrequent patterns could
be used as a subject of research as an alternative since it indicates the absence of frequent patterns. Infrequent pattern
mining is a variation of frequent pattern mining where it finds the uninteresting patterns which rarely occurs. Infrequent
pattern mining has been widely demonstrated its utility in web mining, bioinformatic, medical, genetic and other fields.
Eclat is one of the algorithm which applied in finding frequent patterns in a transaction database. In Eclat variants, iEclat
is the latest algorithm which has a good performance in mining frequent pattern. A few parts of this algorithm need a
modification to assure that it is suitable for mining infrequent pattern. This paper proposes an enhancement algorithm
based on iEclat algorithms for mining infrequent pattern. |
|---|