Approach of Classification Based on Rough Set
Propose a novel approach of classification based on rough set. Fully consider each condition attribute significance and classification object during making policy. It gets higher classification accuracy and less decision rules without attribute reduction by this approach. Experiments have proved its validity.
KeywordsRough set Attribute significance Discretization Classification
This paper is supported by (1) Project of Research on Science and Technology of Jilin Education Ministry of China under Grant No.2007-172 and 2010-383. (2) Science-technology Development Project of Jilin Province of China under Grant No. 20115056. (3) The Natural Science Foundation of Changchun Normal University. (4) The master and doctor launched project of Jilin Normal University.
- 2.Feng L, Wang G, Li T (2009) Knowledge acquisition from decision tables containing continuous-valued attributes [J]. Acta Electron Sinica 37(11):2432–2438Google Scholar
- 3.Wang L, Wu G (2008) Measurements of Discretization Schemes [J]. PR AI 21(4):494–499Google Scholar
- 4.Zhao J, Ni C, Zhan Y, Du Z (2009) Efficient discretizathion algorithm for continuous attributes [J]. Syst Eng Electron 31(1):195–199Google Scholar
- 8.Wang J, Liang J, Qian Y (2008) Uncertainty measure of rough sets based on a knowledge granulation for incomplete information systems. Int J Uncertain Fuzziness Knowl Based Syst 16(2):233–244Google Scholar
- 9.Yue H, Yan D (2010) New algorithm for discretization based on information entropy [J]. Comput Sci 37(4):231–237Google Scholar
- 10.Zhang J, Qiu D (2009) Discretization method based on breakpoint importance in roughness set theory [J]. Meteorol Hydrol Marine Instrum 1:44–47Google Scholar