Empirical Study of Associative Classifiers on Imbalanced Datasets in KEEL
2018
This article presents the empirical performance analysis of the Associative Classification approaches on imbalanced datasets. The imbalanced dataset is a dataset in which ratio of an instance of one class drastically differs from the other one. The ratio difference in class instances, imbalanced dataset, highly affects the performance of the classifiers. An associative Classification is a hybrid technique which combines the classification rules discovery and association rules discovery both are important tasks of Knowledge Discovery. We investigate the performance of selective associative classifiers namely CBA, CBA2, CMAR-C, CPAR-C, and Fuzzy- FARCHD-C by using the methods implemented in KEEL data mining tool on public imbalanced datasets. The experimental results show that the performance of the Fuzzy-FARCHD-C is promising with respect other methods in terms of accuracy.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
19
References
3
Citations
NaN
KQI