Home > Published Issues > 2015 > Volume 6, No. 3, August 2015 >

SUFFUSE: Simultaneous Fuzzy-Rough Feature-Sample Selection

Javad Rahimipour Anaraki 1, Saeed Samet 2, Jeon-Hyun Lee 3, and Chang-Wook Ahn 3
1. Department of Computer Science, Memorial University of Newfoundland, St. John’s, NL, Canada
2. Faculty of Medicine, Memorial University of Newfoundland, St. John’s, NL, Canada
3. Department of Computer Engineering, Sungkyunkwan University, Suwon, South Korea

Abstract—One of the most successful tools for modeling and dealing with uncertainty is Rough Set Theory. Based on this theory several Feature Selection methods have been proposed. As an extension, Fuzzy-Rough set has been introduced to deal with vagueness of both discrete and continuous data in Feature and Sample Selection methods. However, both Fuzzy-Rough Sample Selection and Simultaneous Fuzzy-Rough Feature-Sample Selection are investigated by few. This paper proposes a novel Simultaneous Fuzzy-Rough Feature-Sample Selection method based on Shuffled Frog Leaping Algorithm. The effectiveness of proposed method demonstrated and compared through its performance resulting from nine conventional as well as an improved mGP classifiers over fifteen UCI datasets. This work is also applied to a real world classification problem of noisy Functional Near-Infra-red Spectroscopy neural signals. Experimental results show meaningful increase in classification accuracy, and decrease in dataset size according to non-parametric statistical analysis.

Index Terms—fuzzy-rough sets, simultaneous fuzzy-rough feature-sample selection, feature selection, sample selection

Cite: Javad Rahimipour Anaraki, Saeed Samet, Jeon-Hyun Lee, and Chang-Wook Ahn, "SUFFUSE: Simultaneous Fuzzy-Rough Feature-Sample Selection," Vol. 6, No. 3, pp. 103-110, August, 2015. doi: 10.12720/jait.6.3.103-110