Implicit deep learning prediction rules generalize the recursive rules of
feedforward neural networks. Such rules are based on the solution of a
fixed-point equation involving a single vector of hidden features, which is
thus only implicitly defined. The implicit framework greatly simplifies the
notation of deep learning, and opens up many new possibilities, in terms of
novel architectures and algorithms, robustness analysis and design,
interpretability, sparsity, and network architecture optimization.
%0 Generic
%1 ghaoui2019implicit
%A Ghaoui, Laurent El
%A Gu, Fangda
%A Travacca, Bertrand
%A Askari, Armin
%A Tsai, Alicia Y.
%D 2019
%K deep-implicit-learning from:adulny ode
%T Implicit Deep Learning
%U http://arxiv.org/abs/1908.06315
%X Implicit deep learning prediction rules generalize the recursive rules of
feedforward neural networks. Such rules are based on the solution of a
fixed-point equation involving a single vector of hidden features, which is
thus only implicitly defined. The implicit framework greatly simplifies the
notation of deep learning, and opens up many new possibilities, in terms of
novel architectures and algorithms, robustness analysis and design,
interpretability, sparsity, and network architecture optimization.
@misc{ghaoui2019implicit,
abstract = {Implicit deep learning prediction rules generalize the recursive rules of
feedforward neural networks. Such rules are based on the solution of a
fixed-point equation involving a single vector of hidden features, which is
thus only implicitly defined. The implicit framework greatly simplifies the
notation of deep learning, and opens up many new possibilities, in terms of
novel architectures and algorithms, robustness analysis and design,
interpretability, sparsity, and network architecture optimization.},
added-at = {2021-02-24T10:21:08.000+0100},
author = {Ghaoui, Laurent El and Gu, Fangda and Travacca, Bertrand and Askari, Armin and Tsai, Alicia Y.},
biburl = {https://www.bibsonomy.org/bibtex/21a080e19c280e1646c97b1a7d8364058/adulny},
description = {Implicit Deep Learning - 1908.06315.pdf},
interhash = {fcd13d8b28ba415db1f058c0e05f3261},
intrahash = {1a080e19c280e1646c97b1a7d8364058},
keywords = {deep-implicit-learning from:adulny ode},
note = {cite arxiv:1908.06315},
timestamp = {2021-02-24T10:21:08.000+0100},
title = {Implicit Deep Learning},
url = {http://arxiv.org/abs/1908.06315},
year = 2019
}