Handwritten digit recognition by combined classifiers
Kybernetika, Tome 34 (1998) no. 4, p. [381]
Voir la notice de l'article provenant de la source Czech Digital Mathematics Library
Classifiers can be combined to reduce classification errors. We did experiments on a data set consisting of different sets of features of handwritten digits. Different types of classifiers were trained on these feature sets. The performances of these classifiers and combination rules were tested. The best results were acquired with the mean, median and product combination rules. The product was best for combining linear classifiers, the median for $k$-NN classifiers. Training a classifier on all features did not result in less errors.
@article{KYB_1998__34_4_a4,
author = {van Breukelen, M. and Duin, R. P. W. and Tax, D. M. J. and Hartog, J. E. den},
title = {Handwritten digit recognition by combined classifiers},
journal = {Kybernetika},
pages = {[381]},
publisher = {mathdoc},
volume = {34},
number = {4},
year = {1998},
zbl = {1274.68403},
language = {en},
url = {http://geodesic.mathdoc.fr/item/KYB_1998__34_4_a4/}
}
TY - JOUR AU - van Breukelen, M. AU - Duin, R. P. W. AU - Tax, D. M. J. AU - Hartog, J. E. den TI - Handwritten digit recognition by combined classifiers JO - Kybernetika PY - 1998 SP - [381] VL - 34 IS - 4 PB - mathdoc UR - http://geodesic.mathdoc.fr/item/KYB_1998__34_4_a4/ LA - en ID - KYB_1998__34_4_a4 ER -
van Breukelen, M.; Duin, R. P. W.; Tax, D. M. J.; Hartog, J. E. den. Handwritten digit recognition by combined classifiers. Kybernetika, Tome 34 (1998) no. 4, p. [381]. http://geodesic.mathdoc.fr/item/KYB_1998__34_4_a4/