On the difficulty of training Recurrent Neural Networks
arXiv (Cornell University)2012
Citations Over Time
Abstract
There are two widely known issues with properly training Recurrent Neural Networks, the vanishing and the exploding gradient problems detailed in Bengio et al. (1994). In this paper we attempt to improve the understanding of the underlying issues by exploring these problems from an analytical, a geometric and a dynamical systems perspective. Our analysis is used to justify a simple yet effective solution. We propose a gradient norm clipping strategy to deal with exploding gradients and a soft constraint for the vanishing gradients problem. We validate empirically our hypothesis and proposed solutions in the experimental section.
Related Papers
- → Simple endotrivial modules for quasi-simple groups(2013)26 cited
- → Fundamental relations in simple and 0-simple semihypergroups of small size(2012)8 cited
- → New Characterizations of Simple Points, Minimal Non-simple Sets and P-Simple Points in 2D, 3D and 4D Discrete Spaces(2008)6 cited
- The [0-Simple] Simple Subsemigroups of Nonnegative Matrices(2000)