Lipschitz-bounded 1D convolutional neural networks using the Cayley transform and the controllability Gramian

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Pauli, Patricia, Wang, Ruigang, Manchester, Ian R., Allgöwer, Frank
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929222696566784
author Pauli, Patricia
Wang, Ruigang
Manchester, Ian R.
Allgöwer, Frank
author_facet Pauli, Patricia
Wang, Ruigang
Manchester, Ian R.
Allgöwer, Frank
contents We establish a layer-wise parameterization for 1D convolutional neural networks (CNNs) with built-in end-to-end robustness guarantees. In doing so, we use the Lipschitz constant of the input-output mapping characterized by a CNN as a robustness measure. We base our parameterization on the Cayley transform that parameterizes orthogonal matrices and the controllability Gramian of the state space representation of the convolutional layers. The proposed parameterization by design fulfills linear matrix inequalities that are sufficient for Lipschitz continuity of the CNN, which further enables unconstrained training of Lipschitz-bounded 1D CNNs. Finally, we train Lipschitz-bounded 1D CNNs for the classification of heart arrythmia data and show their improved robustness.
format Preprint
id arxiv_https___arxiv_org_abs_2303_11835
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Lipschitz-bounded 1D convolutional neural networks using the Cayley transform and the controllability Gramian
Pauli, Patricia
Wang, Ruigang
Manchester, Ian R.
Allgöwer, Frank
Machine Learning
Systems and Control
We establish a layer-wise parameterization for 1D convolutional neural networks (CNNs) with built-in end-to-end robustness guarantees. In doing so, we use the Lipschitz constant of the input-output mapping characterized by a CNN as a robustness measure. We base our parameterization on the Cayley transform that parameterizes orthogonal matrices and the controllability Gramian of the state space representation of the convolutional layers. The proposed parameterization by design fulfills linear matrix inequalities that are sufficient for Lipschitz continuity of the CNN, which further enables unconstrained training of Lipschitz-bounded 1D CNNs. Finally, we train Lipschitz-bounded 1D CNNs for the classification of heart arrythmia data and show their improved robustness.
title Lipschitz-bounded 1D convolutional neural networks using the Cayley transform and the controllability Gramian
topic Machine Learning
Systems and Control
url https://arxiv.org/abs/2303.11835