“…Specifically, we have tested five types of classifiers, namely k Nearest Neighbors (k-NN), Linear Discriminant Analysis (LDA), Logistic Regression, Naive Bayes and linear Support Vector Machines (SVMs) [14,4]. Though different approaches motivated these algorithms, all of them are simple, easily adaptable to the multiclass scenario and they have already been used in the electronic nose domain [21,5,59,18,15].…”