Classifier comparison and sensor selection for e-noses

Data derived from chemical vapor sensor arrays was analyzed using Fischer linear discriminant analysis (LDA) and multilayer perceptrons (MLP). Datasets produced by an e-nose based on polymer-carbon black sensors and those produced by a second e-nose based on five thin film metal oxide sensors was analyzed. The datasets were chosen to give classification problems of varying hardness: from the discrimination of two almost Gaussian distributed analytes to the discrimination of two analytes in the presents of interferents at different concentration levels. The performance of the best 5-sensors subset selected with MLP was found to be better than the performance of the LDA-selected subsets.