Comparing Dynamics: Deep Neural Networks versus Glassy Systems

1 code implementation • ICML 2018 • M. Baity-Jesi, L. Sagun, M. Geiger, S. Spigler, G. Ben Arous, C. Cammarota, Y. LeCun, M. Wyart, G. Biroli

We analyze numerically the training dynamics of deep neural networks (DNN) by using methods developed in statistical physics of glassy systems.

Gradient-based learning applied to document recognition

no code implementations • Proceedings of the IEEE 1998 • Y. LeCun, L. Bottou, Y. Bengio, P. Haffner

It uses convolutional neural network character recognizers combined with global training techniques to provide record accuracy on business and personal cheques.

Handwriting Recognition Handwritten Digit Recognition +1