TY - GEN
T1 - Better digit recognition with a committee of simple neural nets
AU - Meier, Ueli
AU - Cireşan, Dan Claudiu
AU - Gambardella, Luca Maria
AU - Schmidhuber, Jürgen
N1 - Generated from Scopus record by KAUST IRTS on 2022-09-14
PY - 2011/12/2
Y1 - 2011/12/2
N2 - We present a new method to train the members of a committee of one-hidden-layer neural nets. Instead of training various nets on subsets of the training data we preprocess the training data for each individual model such that the corresponding errors are decor related. On the MNIST digit recognition benchmark set we obtain a recognition error rate of 0.39%, using a committee of 25 one-hidden-layer neural nets, which is on par with state-of-the-art recognition rates of more complicated systems. © 2011 IEEE.
AB - We present a new method to train the members of a committee of one-hidden-layer neural nets. Instead of training various nets on subsets of the training data we preprocess the training data for each individual model such that the corresponding errors are decor related. On the MNIST digit recognition benchmark set we obtain a recognition error rate of 0.39%, using a committee of 25 one-hidden-layer neural nets, which is on par with state-of-the-art recognition rates of more complicated systems. © 2011 IEEE.
UR - http://ieeexplore.ieee.org/document/6065510/
UR - http://www.scopus.com/inward/record.url?scp=82355173010&partnerID=8YFLogxK
U2 - 10.1109/ICDAR.2011.252
DO - 10.1109/ICDAR.2011.252
M3 - Conference contribution
SN - 9780769545202
SP - 1250
EP - 1254
BT - Proceedings of the International Conference on Document Analysis and Recognition, ICDAR
ER -