Neural networks for odor recognition in artificial noses

This paper compares the results obtained by different neural network approaches for two odor recognition problems. The data acquisitions were performed by different prototypes of artificial noses. The aim is to provide a comparative analysis of different neural network models, such as multi-layer perceptrons, time delay neural networks and radial basis function networks, for the classification of odors from different substances in artificial noses. Simulated annealing and tabu search were used to optimize neural networks.