Neural Tree Network Based Electronic Nose
The training of a multi-layer perceptron using the well known back-propagation algorithm usually requires a priori knowledge of the network architecture. In this paper, results are presented on two practical classification problems which use neural tree classifiers to determine automatically the optimal number of neurons required. The data-set comes from the response of the Warwick Electronic Nose to a set of simple and complex odours.
KeywordsInternal Node Electronic Nose Decision Tree Classifier Momentum Coefficient Input Feature Vector
Unable to display preview. Download preview PDF.
- Gardner J W, Bartlett P N, Dodd G H and Shurmer H V, ‘Chemosensory information processing’, ed. D Schild (Berlin: Springer) Vol H39, pp 131, 1990.Google Scholar
- Rumelhart D E and McClelland J L, ‘Parallel distributed processing’, MIT press, chapter 8, 1986.Google Scholar
- Lipmann R P, ‘An introduction to computing with neural nets’, IEEE ASSP Magazine, pp 4-22, April 1987.Google Scholar
- Brent R P, ‘Fast Training Algorithms for Multilayer Neural Nets’, IEEE Trans. on Neural Networks, pp 346-354, 1991.Google Scholar
- Sirat J A and Nadal J-P, ‘Neural trees: a new tool for classification’, Network, 1990.Google Scholar
- Sankar A and Mammone R J, ‘Speaker Independent Vowel Recognition using Neural Tree Networks’.Google Scholar
- Mammone R J and Zeevi Y Y, ‘Neural Networks: theory and application’, Academic Press, 1991.Google Scholar
- Hines E L, Gardner J W, Fung W and Fekadu A A, ‘Improved rate of convergence in a MLP based electronic nose’, 2nd Irish Neural Networks Conf., Belfast, June 1992.Google Scholar
- Leung C W, ‘Application of neural networks to the classification of coffee data’, MSc IT Dissertation, Department of Engineering, University of Warwick, 1991.Google Scholar