doi: 10.18178/ijfcc.2017.6.3.494
A Method to Clustering the Feature Ranking on Data Classification Using an Ensemble Feature Selection
Nuntawut Kaoungku
1
,
Kittisak Kerdprasop
1,2
,
Nittaya Kerdprasop
1,3
- 1 School of Computer Engineering, Suranaree University of Technology, Nakhon Ratchasima 30000, Thailand
- 2 Knowledge Engineering Research Unit, Suranaree University of Technology, Nakhon Ratchasima 30000, Thailand
- 3 Data Engineering Research Unit, Suranaree University of Technology, Nakhon Ratchasima 30000, Thailand
Abstract
The aim of this paper is to improve the predictiveperformance of the classification process by means of buildingmultiple data classification models based on the output fromfeature selection methods that use ensemble strategy to find theoptimal set of features. Currently, the data volume has grown atan extreme rate causing a variety of problems. The big datasituation has made automatic analysis tasks such as dataclassification facing low performance and high computationaltime problems when dealing with big data that are huge in bothvolume and dimensions. In this research work, we tackle the bigdata problem in the high dimensionality aspect. We propose anensemble method to reduce data dimension by means of featureclustering to rank the potential features and also return suitablesubset of features for further classifying the training data. Thetwo paradigms of feature selection based on ensemble strategyare proposed and evaluated. Experimental results confirm theefficacy of our proposed feature ensemble method.
Keywords
- Feature selection
- ensemble learning
- clustering
- classification
How to Cite
Nuntawut Kaoungku, Kittisak Kerdprasop, and Nittaya Kerdprasop, "A Method to Clustering the Feature Ranking on Data Classification Using an Ensemble Feature Selection," International Journal of Future Computer and Communication, vol. 6, no. 3, pp. 81-85, 2017. https://doi.org/ 10.18178/ijfcc.2017.6.3.494
Copyright & License
Copyright © 2017 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).