Qingfeng Duan
School of Economic and Management, North University of China, Taiyuan, Shanxi, 030051, China
ABSTRACT
Decision-theoretic rough set model is applied to the feature selection with capability of error tolerance which could deal with the decision problem with missing value or noise. Thus, a novel cost sensitive approach of feature selection is proposed. It tends to find appropriate reduct by the criterion of minimum cost computed by the theory of three-way Bayesian decisions and takes a reasonable selection with a heuristic strategy based on mutual information theory. To validate the proposed approach, classification performance comparison is employed on eight UCI datasets empirically. Consequently, it is demonstrated that the proposed approach work well and outperforms the others chosen in the experiments at most cases.
PDF References Citation
How to cite this article
Qingfeng Duan, 2013. A Novel Approach of Feature Selection Based on Decision-theoretic Rough Set Model. Information Technology Journal, 12: 5226-5230.
DOI: 10.3923/itj.2013.5226.5230
URL: https://scialert.net/abstract/?doi=itj.2013.5226.5230
DOI: 10.3923/itj.2013.5226.5230
URL: https://scialert.net/abstract/?doi=itj.2013.5226.5230
REFERENCES
- Chen, Y., D. Miao, R. Wang and K. Wu, 2011. A rough set approach to feature selection based on power set tree. Knowledge-Based Syst., 24: 275-281.
CrossRef - Derrac, J., C. Cornelis, S. Garcia and F. Herrera, 2012. Enhancing evolutionary instance selection algorithms by means of fuzzy rough set based feature selection. Inform. Sci., 186: 73-92.
CrossRef - Inuiguchi, M., Y. Yoshioka and Y. Kusunoki, 2009. Variable-precision Dominance-based rough set approach and attribute reduction. Int. J. Approx. Reason., 50: 1199-1214.
CrossRef - Jia, X., W. Liao, Z. Tang and L. Shang, 2013. Minimum cost attribute reduction in decision-theoretic rough set models. Inform. Sci., 219: 151-167.
CrossRef - Liang, J., F. Wang, C. Dang and Y. Qian, 2012. An efficient rough feature selection algorithm with a multi-granulation view. Int. J. Approximate Reasoning, 53: 912-926.
CrossRef - Miao, D. and L. Hou, 2004. A comparison of rough set methods and representative inductive learning algorithms. Fundamenta Inform., 59: 203-219.
Direct Link - Qian, Y., J. Liang, W. Pedrycz and C. Dang, 2011. An efficient accelerator for attribute reduction from incomplete data in rough set framework. Pattern Recognit., 44: 1658-1670.
CrossRef - Salamo, M. and M. Lopez-Sanchez, 2011. Rough set based approaches to feature selection for Case-based reasoning classifiers. Pattern Recognit. Lett., 32: 280-292.
CrossRef - Yao, Y., 2010. Three-way decisions with probabilistic rough sets. Inform. Sci., 180: 341-353.
CrossRef