Convex neural network synthesis for robustness in the 1-norm

Ross Drummond, Chris Guiver, Matthew Turner
Proceedings of the 6th Annual Learning for Dynamics & Control Conference, PMLR 242:1388-1399, 2024.

Abstract

With neural networks being used to control safety-critical systems, they increasingly have to be both accurate (in the sense of matching inputs to outputs) and robust. However, these two properties are often at odds with each other and a trade-off has to be navigated. To address this issue, this paper proposes a method to generate an approximation of a neural network which is certifiably more robust. Crucially, the method is fully convex and posed as a semi-definite programme. An application to robustifying model predictive control is used to demonstrate the results. The aim of this work is to introduce a method to navigate the neural network robustness/accuracy trade-off.

Cite this Paper

BibTeX
@InProceedings{pmlr-v242-drummond24b, title = {Convex neural network synthesis for robustness in the 1-norm}, author = {Drummond, Ross and Guiver, Chris and Turner, Matthew}, booktitle = {Proceedings of the 6th Annual Learning for Dynamics & Control Conference}, pages = {1388--1399}, year = {2024}, editor = {Abate, Alessandro and Cannon, Mark and Margellos, Kostas and Papachristodoulou, Antonis}, volume = {242}, series = {Proceedings of Machine Learning Research}, month = {15--17 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v242/drummond24b/drummond24b.pdf}, url = {https://proceedings.mlr.press/v242/drummond24b.html}, abstract = {With neural networks being used to control safety-critical systems, they increasingly have to be both accurate (in the sense of matching inputs to outputs) and robust. However, these two properties are often at odds with each other and a trade-off has to be navigated. To address this issue, this paper proposes a method to generate an approximation of a neural network which is certifiably more robust. Crucially, the method is fully convex and posed as a semi-definite programme. An application to robustifying model predictive control is used to demonstrate the results. The aim of this work is to introduce a method to navigate the neural network robustness/accuracy trade-off.} }
Endnote
%0 Conference Paper %T Convex neural network synthesis for robustness in the 1-norm %A Ross Drummond %A Chris Guiver %A Matthew Turner %B Proceedings of the 6th Annual Learning for Dynamics & Control Conference %C Proceedings of Machine Learning Research %D 2024 %E Alessandro Abate %E Mark Cannon %E Kostas Margellos %E Antonis Papachristodoulou %F pmlr-v242-drummond24b %I PMLR %P 1388--1399 %U https://proceedings.mlr.press/v242/drummond24b.html %V 242 %X With neural networks being used to control safety-critical systems, they increasingly have to be both accurate (in the sense of matching inputs to outputs) and robust. However, these two properties are often at odds with each other and a trade-off has to be navigated. To address this issue, this paper proposes a method to generate an approximation of a neural network which is certifiably more robust. Crucially, the method is fully convex and posed as a semi-definite programme. An application to robustifying model predictive control is used to demonstrate the results. The aim of this work is to introduce a method to navigate the neural network robustness/accuracy trade-off.
APA
Drummond, R., Guiver, C. & Turner, M.. (2024). Convex neural network synthesis for robustness in the 1-norm. Proceedings of the 6th Annual Learning for Dynamics & Control Conference, in Proceedings of Machine Learning Research 242:1388-1399 Available from https://proceedings.mlr.press/v242/drummond24b.html.

Related Material