English

Better Training using Weight-Constrained Stochastic Dynamics

Machine Learning 2021-06-22 v1 Machine Learning

Abstract

We employ constraints to control the parameter space of deep neural networks throughout training. The use of customized, appropriately designed constraints can reduce the vanishing/exploding gradients problem, improve smoothness of classification boundaries, control weight magnitudes and stabilize deep neural networks, and thus enhance the robustness of training algorithms and the generalization capabilities of neural networks. We provide a general approach to efficiently incorporate constraints into a stochastic gradient Langevin framework, allowing enhanced exploration of the loss landscape. We also present specific examples of constrained training methods motivated by orthogonality preservation for weight matrices and explicit weight normalizations. Discretization schemes are provided both for the overdamped formulation of Langevin dynamics and the underdamped form, in which momenta further improve sampling efficiency. These optimization schemes can be used directly, without needing to adapt neural network architecture design choices or to modify the objective with regularization terms, and see performance improvements in classification tasks.

Keywords

Cite

@article{arxiv.2106.10704,
  title  = {Better Training using Weight-Constrained Stochastic Dynamics},
  author = {Benedict Leimkuhler and Tiffany Vlaar and Timothée Pouchon and Amos Storkey},
  journal= {arXiv preprint arXiv:2106.10704},
  year   = {2021}
}

Comments

ICML 2021 camera-ready. arXiv admin note: substantial text overlap with arXiv:2006.10114

R2 v1 2026-06-24T03:24:03.359Z