Clustering algorithm based on k-means and fuzzy entropy for e-nose applications

Abstract

Clustering is the major research area in the field of pattern recognition especially for artificial sensing systems like e-nose applications. The main goal of this paper is to develop a fuzzy clustering algorithm having application for classifying electronic nose data. In this paper, a two step clustering algorithm is proposed. In first step k-means algorithm of clustering was applied on each data dimension of data under investigation and in next step, fuzzy entropy of each dimension was calculated. The fuzzy entropy is calculated on membership value of the data points. Labeling of final data class was performed on the basis of fuzzy entropy, which improves accuracy of the traditional k-means algorithm. Finally, the proposed algorithm has been tested on experimental data set of electronic nose.