Fuzzy rough set approach based classifier

Authors:
Alpna Singh;Aruna Tiwari;Sujata Naegi
Affiliations:
Department of Computer Engineering, SGSITS Indore, India;Department of Computer Engineering, SGSITS Indore, India;Department of Computer Engineering, SGSITS Indore, India
Venue:
SEMCCO'11 Proceedings of the Second international conference on Swarm, Evolutionary, and Memetic Computing - Volume Part I
Year:
2011

Citing 6
Cited 0

Learning Classification Rules from Database in the Context of Knowledge Acquisition and Representation

IEEE Transactions on Knowledge and Data Engineering
Generalized fuzzy rough sets

Information Sciences—Informatics and Computer Science: An International Journal
Constructive and axiomatic approaches of fuzzy approximation operators

Information Sciences—Informatics and Computer Science: An International Journal - Mining stream data
Learning fuzzy rules from fuzzy samples based on rough set technique

Information Sciences: an International Journal
Building a Rule-Based Classifier—A Fuzzy-Rough Set Approach

IEEE Transactions on Knowledge and Data Engineering
On the generalization of fuzzy rough sets

IEEE Transactions on Fuzzy Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper a fuzzy rough set approach based classifier is proposed. To design this classifier, fuzzy approximation operator proposed by Zhao, Tsang and Chen 2010 [1] has been modified and new rules are proposed for classification. The fuzzy rough set based classification rules are used to predict decision class of new objects with unknown class. To extract these rules, first we build the equivalence classes, calculate the lower approximation value and then make use of a constant degree to reduce redundant attribute values. By using this concept, we design the discernibility vector, attribute value core of every object, to develop an attribute value reduction algorithm. These rules are applied on benchmark of dataset and classification accuracy is measured. Experimental results have been carried out and it shows that number of rules, training time of classifier is reduced and classification accuracy is improved on some dataset.