Abstract
It has been established that committee classifiers, in which the outputs of different, individual network classifiers are combined in various ways, can produce better accuracy than the best individual in the committee. We describe results showing that these advantages are obtained when neural networks are applied to a taxonomic problem in marine science: the classification of images of marine phytoplankton. Significant benefits were found when individual networks, trained on different classes of input, having comparable individual performances, were combined. Combining networks of very different accuracy did not improve performance when measured against the best single network, but nor was it reduced. An alternative architecture, which we term a collective machine, in which the different data types are combined in a single network, was found to have significantly better accuracy than the committee machine architectures. The performance gains and resilience to non-discriminatory types of data suggest the techniques have great utility in the development of general purpose, network classifiers.
Similar content being viewed by others
References
Gaston KJ, May RM. Taxonomy of taxonomists. Nature 1992; 356: 281–282
Sokal RR. Classification: purposes, principles, progress, prospects. Science 1974; 185 (27): 1115–1123
Simpson R, Williams R, Ellis R, Culverhouse PF. Biological pattern recognition by neural networks. Marine Ecology Series 1992; 79: 303–308
Caiti A, Magenes G, Parisini T, Simpson R. Smooth approximations by RBFs: three case studies. J Applied Sciences and Computations 1994; 1: 88–112
Ellis R, Simpson R, Culverhouse PF, Parisini T, Williams R, Requera B, Moore M, Lowe D. Expert visual classification and neural networks: can general solutions be found? Proc Oceans '94, 1994, 330–334
Simpson R, Culverhouse PF, Parisini T, Ellis R, Lindley AJ, Williams R, Requera B. Automatic neural network classification of 23 species of phytoplankton. Technical Report, CT92-0015-22, University of Plymouth, 1992
Jacobs RA, Jordan MI, Nowlan SJ, Hinton GE. Adaptive mixtures of local experts. Neural Computation 1991; 3(1): 79–87
Wolpert DH. Stacked generalisations. Neural Networks 1992; 5: 241–259.
Battiti R, Colla AM. Democracy in neural nets: voting schemes for classification. Neural Networks 1994; 7 (4): 691–707
Gabor D. Theory of communication. J Institute of Electrical Engineers 1946; 93: 429–457
Daugman JG. An information-theoretic view of analogue representation in striate cortex. In Schwartz EL. (ed), Computational Neuroscience, MIT Press, Cambridge, MA, 1990, 403–24
Bovik AC, Clark M, Geisler WE. Multichannel texture analysis using localised spatial filters. IEEE Pattern Analysis and Machine Intelligence 1990; 12: 55–73
Ballard DH, Brown CM. Computer Vision, Prentice-Hall, Englewood Cliffs, NJ, 1982.
Ramsay A, Barrett R. AI in Practice: Examples in POP-11, Ellis Horwood, Chichester, 1987, 191–197
Rumelhart DE, Hinton GE, Williams RJ. Learning internal representations by error propagation. In Rumelhart DE, McClelland JL. (eds.), Parallel Distributed Processing: Explorations in the microstructure of cognition, MIT Press, Cambridge, MA, 1986, 318–364
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Ellis, R., Simpson, R., Culverhouse, P.F. et al. Committees, collectives and individuals: Expert visual classification by neural network. Neural Comput & Applic 5, 99–105 (1997). https://doi.org/10.1007/BF01501174
Issue Date:
DOI: https://doi.org/10.1007/BF01501174