Greedy Layer-Wise Training of Deep Networks

par/by Yoshua Bengio
Département d'Informatique et de Recherche Opérationnelle
Université de Montréal

Deep multi-layer neural networks have many levels of non-linearities allowing them to compactly represent highly non-linear and highly-varying functions. However, until recently it was not clear how to train such deep networks, since gradientbased optimization starting from random initialization appears to often get stuck in poor solutions. Hinton et al. recently introduced a greedy layer-wise unsupervised learning algorithmfor Deep Belief Networks (DBN), a generative model with many layers of hidden causal variables.

In the context of the above optimization problem, we present an empirical study of this algorithm and explore variants to better understand its success and extend it to cases where the inputs are continuous or where the structure of the input distribution is not revealing enough about the variable to be predicted in a supervised task. Our experiments also confirm the hypothesis that the greedy layer-wise unsupervised training strategy mostly helps the optimization, by initializing weights in a region near a good local minimum, giving rise to internal distributed representations that are high-level abstractions of the input bringing better generalization.