Prototypes Reduction and Feature Selection based on Fuzzy Boundary Area for Nearest Neighbor Classifiers
- 20 Downloads
For prototype-based classifiers, the number of prototypes results in increasing the computational time so that it takes very long time for a prototype-based classifier to determine the class label of an associated data. Many researchers have been interested in the reduction of the number of prototypes without degradation of the classification ability of prototype-based classifiers. In this paper, we introduce a new method for generating prototypes based on the assumption that the prototypes positioned near the boundary surface are important for improving the classification abilities of nearest neighbor classifiers. The main issue of this paper is how to locate the new prototypes as close as possible to the boundary surface. To realize this, we consider possibilistic C-Means clustering and conditional C-Means clustering. The clusters obtained by using possibilistic C-Means clustering methods are used to define the boundary areas, and the conditional fuzzy C-Means clustering technique is used to determine the locations of prototypes within the already defined boundary areas. The design procedure is illustrated with the aid of numeric examples that provide a thorough insight into the effectiveness of the proposed method.
KeywordsPrototype-based classifier Possibilistic C-means clustering Prototype reduction Conditional C-means clustering Boundary area definition Nearest neighbor classifier
This paper was supported by Wonkwang University in 2017.
- 4.Dasarathy, B.V.: Nearest Neighbor (NN) Norms: NN Pattern Classification Techniques. IEEE Computer Society Press (1991)Google Scholar
- 5.Maron, O., Ratan, A.L.: Multiple-instance learning for natural scene classification. In: Proceedings of the 15th International Conference On Machine Learning, Morgan Kaufmann, San Francisco, pp. 341–349 (1998)Google Scholar
- 6.Kim, D.-E., Yu, J.-H., & Sim, K.-B.: EEG feature classification based on grip strength for BCI applications, International Journal of Fuzzy Logic and Intelligent Systems, 15(4), 277-282 (2015)Google Scholar
- 15.Pedrycz, W.: Conditional fuzzy C-means. Pattern Recogn. Lett. 17(6), 626–632 (1996)Google Scholar
- 21.Devijver, P.A., Kittler, J.: On the edited nearest neighbor rule. In: Proceedings of the Fifth International Conference on Pattern Recognition, IEEE Computer Society, Miami, pp. 72–80 (1984)Google Scholar
- 35.Lee, S.Y., Urtnasan, E., & Lee, K.-J.: Design of a fast learning classifier for sleep apnea database based on fuzzy SVM, International Journal of Fuzzy Logic and Intelligent Systems, 17(3), 187-193 (2017)Google Scholar