Pattern Classification Using Class-Dependent Rough-Fuzzy Granular Space
The article describes a new rough-fuzzy model for pattern classification. Here, class-dependent granules are formulated in fuzzy environment that preserve better class discriminatory information. Neighborhood rough sets (NRS) are used in the selection of a subset of granulated features that explore the local/contextual information from neighbor granules. The model thus explores mutually the advantages of class-dependent fuzzy granulation and NRS that is useful in pattern classification with overlapping classes. The superiority of the proposed model to other similar methods is demonstrated with both completely and partially labeled data sets using various performance measures. The proposed model learns well even with a lower percentage of training set that makes the system fast.
KeywordsFuzzy information granulation rough neighborhood sets rough-fuzzy granular computing pattern recognition soft computing
Unable to display preview. Download preview PDF.
- 2.Zadeh, L.A.: Fuzzy sets and information granularity. In: Gupta, M., Ragade, R., Yager, R. (eds.) Advances in Fuzzy Set Theory and Applications, pp. 3–18. North-Holland Publishing Co., Amsterdam (1979)Google Scholar
- 7.Lin, T.Y.: Granulation and nearest neighborhoods: rough set approach. In: Pedrycz, W. (ed.) Granular computing: an emerging paradigm, pp. 125–142. Physica-Verlag, Heidelberg (2001)Google Scholar
- 14.Elena database, ftp://ftp.dice.ucl.ac.be/pub/neural-nets/ELENA/databases/REAL/
- 15.NRSA, IRS data users hand book, Technical report, Document No. IRS/NRSA/NDC/HB-02/89 (1989)Google Scholar