Abstract
This paper proposes the use of interpolating vectors for robust pattern recognition. Labeled reference vectors in a multi-dimensional feature space are first produced by a kind of competitive learning. We then assume a situation where interpolating vectors are densely placed along lines connecting all pairs of reference vectors of the same label. From these interpolating vectors, we choose the one that has the largest similarity to the test vector. Its label shows the result of pattern recognition. We applied this method to the neocognitron for handwritten digit recognition, and reduced the error rate from 1.48% to 1.00% for a blind test set of 5000 digits.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Cover, T.M., Hart, P.E.: Nearest neighbor pattern classification. IEEE Transactions on Information Theory IT-4, 515–516 (1968)
Gray, R.M.: Vector quantization. IEEE ASSP Magazine 1, 4–29 (1984)
Kohonen, K.: The Self-Organizing Maps. Springer, Heidelberg (1995)
Schölkoph, B., Sung, K., Burges, C., Girosi, F., Niyogi, P., Poggio, T., Vapnik, V.: Comparing support vector machines with Gaussian kernels to radial basis function classifiers. IEEE Trans. Signal Processing 45, 2758–2764 (1997)
Fukushima, K.: Neocognitron: a hierarchical neural network capable of visual pattern recognition. Neural Networks 1, 119–130 (1988)
Fukushima, K.: Neocognitron for handwritten digit recognition. Neurocomputing 51, 161–180 (2003); A computer program of this neocognitron in C language is available from Visiome Platform: http://platform.visiome.neuroinf.jp/
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 2008 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Fukushima, K. (2008). Interpolating Vectors: Powerful Algorithm for Pattern Recognition. In: Ishikawa, M., Doya, K., Miyamoto, H., Yamakawa, T. (eds) Neural Information Processing. ICONIP 2007. Lecture Notes in Computer Science, vol 4985. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-69162-4_1
Download citation
DOI: https://doi.org/10.1007/978-3-540-69162-4_1
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-69159-4
Online ISBN: 978-3-540-69162-4
eBook Packages: Computer ScienceComputer Science (R0)