Deep, Big, Simple Neural Nets for Handwritten Digit Recognition
Dalle Molle Institute for Artificial Intelligence Research · Università della Svizzera italiana
Abstract
Good old online backpropagation for plain multilayer perceptrons yields a very low 0.35% error rate on the MNIST handwritten digits benchmark. All we need to achieve this best result so far are many hidden layers, many neurons per layer, numerous deformed training images to avoid overfitting, and graphics cards to greatly speed up learning.
Citation impact
- FWCI
- 32.28
- Percentile
- 100%
- References
- 31
Authors
4- DCDan Claudiu CireşanCorresponding
Dalle Molle Institute for Artificial Intelligence Research, Università della Svizzera italiana
- UMUeli Meier
Dalle Molle Institute for Artificial Intelligence Research, Università della Svizzera italiana
- LMLuca Maria Gambardella
Dalle Molle Institute for Artificial Intelligence Research, Università della Svizzera italiana
- JSJürgen Schmidhuber
Dalle Molle Institute for Artificial Intelligence Research, Università della Svizzera italiana
Topics & keywords
- Simple (philosophy)
- Computer science
- Digit recognition
- Artificial neural network
- Pattern recognition (psychology)
- Numerical digit
- Speech recognition
- Artificial intelligence