Development of Fuzzy Learning Vector Quantization Neural Network for Artificial Odor Discrimination System
The author had developed an artificial odor discrimination system for mimicking a function of human odor experts. The system used a back-propagation neural network and shows high recognition capability, however, the system work efficiently if it is used to discriminate a limited number of odors. The back-propagation learning algorithm will force the unlearned odor to the one of the already learned class-category. To improve the system’s capability, a fuzzy learning vector quantization (FLVQ) neural network is developed, in which LVQ neural network will be used together with fuzzy theory. In the experiments on four different ethanol concentrations and three different kinds of fragrance odor from Martha Tilaar Cosmetics, it is found that the FLVQ shows high recognition capability, comparable with the back propagation neural network, however, the FLVQ can cluster the unlearned sample to different class of odor.
KeywordsVector Quantization Back Propagation Neural Network Reference Vector Learning Vector Quantization Recognition Probability
Unable to display preview. Download preview PDF.
- B. Kusumoputro and M. Rivai, ‘Discrimination of fragrance odor by arrayed quartz resonator and a neural network’, ICCIMA-98, H. Selvaraj and B. Verma (Eds), Singapore: Word Scientific, pp. 264–270, 1998.Google Scholar
- T. Kohonen, G. Barna and R. Chrisley, ‘Statistical pattern recognition with neural networks: Benchmarking studies’, IEEE Proc. of ICNN, pp.61–68, 1987.Google Scholar
- T. Kohonen, ‘Improved versions of learning vector quantization’, IEEE Proc. of IJCNN, I, pp. 545–550, 1990.Google Scholar
- W.H. King, ‘Piezoelectric Sorption detector’, Anal. Chem., pp. 206–222, 1964.Google Scholar
- D.E. Rumelhart, J.L. Mc.Cleland, and The PDP Research Group, Parallel Distributed Processing, MIT Press, 1996Google Scholar