ICML 2017poster305 citations
Equivariance Through Parameter-Sharing
Siamak Ravanbakhsh, Jeff Schneider, Barnabás Póczos
Abstract
We propose to study equivariance in deep neural networks through parameter symmetries. In particular, given a group G that acts discretely on the input and output of a standard neural network layer, we show that its equivariance is linked to the symmetry group of network parameters. We then propose two parameter-sharing scheme to induce the desirable symmetry on the parameters of the neural network. Under some conditions on the action of G, our procedure for tying the parameters achieves G-equivariance and guarantees sensitivity to all other permutation groups outside of G.
BibTeX
@InProceedings{pmlr-v70-ravanbakhsh17a,
title = {Equivariance Through Parameter-Sharing},
author = {Siamak Ravanbakhsh and Jeff Schneider and Barnab{\'a}s P{\'o}czos},
booktitle = {Proceedings of the 34th International Conference on Machine Learning},
pages = {2892--2901},
year = {2017},
editor = {Precup, Doina and Teh, Yee Whye},
volume = {70},
series = {Proceedings of Machine Learning Research},
month = {06--11 Aug},
publisher = {PMLR},
pdf = {http://proceedings.mlr.press/v70/ravanbakhsh17a/ravanbakhsh17a.pdf},
url = {https://proceedings.mlr.press/v70/ravanbakhsh17a.html},
abstract = {We propose to study equivariance in deep neural networks through parameter symmetries. In particular, given a group G that acts discretely on the input and output of a standard neural network layer, we show that its equivariance is linked to the symmetry group of network parameters. We then propose two parameter-sharing scheme to induce the desirable symmetry on the parameters of the neural network. Under some conditions on the action of G, our procedure for tying the parameters achieves G-equivariance and guarantees sensitivity to all other permutation groups outside of G.}
}