Paper Info
Title
Fuzzy Relative Entropy Based Classification Scheme For Discrimination Of Odors/Gases Using A Poorly Selective Sensor Array
Abstract
A novel fuzzy classification scheme is presented for discrimination of odours/gases using response of an oxygen plasma treated tin oxide sensor array, capable of sensing at room temperature. The proposed technique employs 'Fuzzy Relative Entropy' as a measure of distinctiveness of an odour sample to classify individual sensor response samples to one of four gas classes. Fuzzy relative entropy is used to first select a standard set of samples from a fuzzified data set and then remaining sets of samples are compared with it for classification. The proposed scheme can be an alternative to popular neural network based classifiers which have issues with convergence. The technique may find application in a plethora of real time odour/gas monitoring and discrimination systems.
Year
Venue
Keywords
2016
2016 IEEE INTERNATIONAL CONFERENCE ON FUZZY SYSTEMS (FUZZ-IEEE)
Electronic nose, Fuzzy relative entropy, Fuzzy pattern matching
Field
DocType
ISSN
Convergence (routing),Data mining,Fuzzy classification,Computer science,Sensor array,Fuzzy set,Artificial intelligence,Artificial neural network,Pattern recognition,Classification scheme,Fuzzy logic,Machine learning,Kullback–Leibler divergence
Conference
1544-5615
Citations 
PageRank 
References 
0
0.34
3
Authors
2
Name
Order
Citations
PageRank
Shilpa Sharma100.34
Ravi Kumar212.04