Classification reliability and its use in multi-classifier systems
In the last years, great attention has been devoted to multiple classifier systems. The implementation of such a system implies the definition of a rule (combining rule) for determining the most likely class, on the basis of the class attributed by each single expert. The availability of a criterion to evaluate the credibility of the decision taken by a classifier can be profitable in order to implement the combining rule. We propose a method that, after defining the reliability of a classification on the basis of information directly derived from the output of the classifier, uses this information in the context of a combining rule. The results obtained by combining four handwritten character on the basis of classification reliability are compared with those obtained by using three different combining criteria. Tests have been performed using a standard handwritten character database.
Unable to display preview. Download preview PDF.
- L. Xu, A. Krzyzak, C.Y. Suen, “Method of Combining Multiple Classifiers and Their Application to Handwritten Numeral Recognition”, IEEE Trans. on Systems, Man and Cybernetics, vol. 22, no. 3, pp. 418–435, January 1992.Google Scholar
- S.-B. Cho, J.H. Kim, “Combining Multiple Neural Networks by Fuzzy Integral for Robust Classification”, IEEE Trans. on Systems, Man and Cybernetics, vol. 25, no. 2, pp. 380–384, February 1995.Google Scholar
- P.J. Grother, NIST Special Database 19, Technical Report, National Institute of Standards and Technology, 1995Google Scholar
- L.P. Cordella, C. De Stefano, M. Vento, “A Neural Network Classifier for OCR using Structural Descriptions”, Machine Vision and Applications, no. 8, pp. 336–342, 1995.Google Scholar
- P. Foggia, C. Sansone, F. Tortorella, and M. Vento, “Character Recognition by Geometrical Moments on Structural Decompositions”, Proc. 4th Int. Conf. on Document Analysis and Recognition, to appear.Google Scholar
- D.E. Rumelhart, J.L. Mc Clelland, Parallel Distributed Processing-Explorations in the Microstructure of Cognition, Vol.1: Foundations. MIT Press, Cambridge, Mass, 1986.Google Scholar
- T. Kohonen, “The Self-Organizing Map,” Proc. of the IEEE, vol. 78, no. 9, pp. 1464–1480, September 1990.Google Scholar
- R. Hecth-Nielsen, Neurocomputing, Addison-Wesley, Reading (MA), 1990.Google Scholar
- C. De Stefano, P. Foggia, F. Tortorella, M. Vento, “A Distance Measure for Structural Descriptions using Circle Arcs as Primitives” in Proc. 13th Int. Conf. on Pattern Recogn., IEEE Comp. Soc. Press, Vol. II, pp. 290–294, 1996.Google Scholar
- L.P. Cordella, C. De Stefano, F. Tortorella, M. Vento, “A Method for Improving Classification Reliability of Multilayer Perceptrons”, IEEE Trans. on Neural Networks, vol. 6, no. 5, pp. 1140–1147, September 1995.Google Scholar
- J. Gordon, E.H. Shortliffe, “The Dempster-Shafer Theory of Evidence”, in B.G. Buchanan, E.H. Shortliffe (Eds.), Rule-Based Expert Systems, Addison-Wesley, pp. 272–292, 1984.Google Scholar