Comparison between different neural network architectures for odour discrimination
Comparison between different neural network architectures for odour discrimination
- Author(s): G. Gestri and A. Starita
- DOI: 10.1049/cp:19950591
For access to this article, please select a purchase option:
Buy conference paper PDF
Buy Knowledge Pack
IET members benefit from discounts to all IET publications and free access to E&T Magazine. If you are an IET member, log in to your account and the discounts will automatically be applied.
4th International Conference on Artificial Neural Networks — Recommend this title to your library
Thank you
Your recommendation has been sent to your librarian.
- Author(s): G. Gestri and A. Starita Source: 4th International Conference on Artificial Neural Networks, 1995 p. 410 – 414
- Conference: 4th International Conference on Artificial Neural Networks
- DOI: 10.1049/cp:19950591
- ISBN: 0 85296 641 5
- Location: Cambridge, UK
- Conference date: 26-28 June 1995
- Format: PDF
With the attempt to develop an artificial olfactory system able to mimic the discrimination ability of the natural system, several artificial neural network architectures were considered and evaluated on the basis of their performances and similarities with the neurophysiological models of the biological system. The neural network architectures were analysed and tested with experimental data from an array of broadly tuned polymers gas sensors. Since these sensors are weakly selective, like the receptors of the natural system, the task of the networks is to obtain the selectivity enhancement that is needed to correctly discriminate the odours. The advantage of using ANN to classify data rather than statistical methods is that doesn't require many assumptions about the form of the data. In addition, ANN can cope with highly non linear data and can be made to cope with noisy or drifting data. The paradigms considered are Counterpropagation Networks (R. Hecht Nielsen, 1987), Bi directional Associative Memories (B. Kosko, 1988), Hamming (R.P. Lippmann, 1987), and Adaptive Resonance Theory Networks (G.A. Carpenter and S. Grossberg, 1988). The results were compared and in a few cases some modifications of the used paradigms have been done to optimise the answer of the system.
Inspec keywords: self-organising feature maps; pattern classification; content-addressable storage; neural net architecture; ART neural nets; gas sensors; chemioception
Subjects: Transducers and sensing devices; Parallel architecture; Signal processing and detection; Digital signal processing; Neural computing techniques; Chemical sensors; Associative storage; Pattern recognition
Related content
content/conferences/10.1049/cp_19950591
pub_keyword,iet_inspecKeyword,pub_concept
6
6