ABBA neural networks: coping with positivity, expressivity, and robustness

DOI10.1137/23M1589591MaRDI QIDQ6583716zbMATH OpenFDO

Authors Ana Neacşu, Jean-Christophe Pesquet, Vlad Vasilescu, Corneliu Burileanu

Publication date 6 August 2024

Published in SIAM Journal on Mathematics of Data Science (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1137/23M1589591

zbMATH Keywords

image classificationLipschitzneural networksoptimizationrobustness

Mathematics Subject Classification ID

Numerical optimization and variational techniques (65K10) Learning and adaptive systems in artificial intelligence (68T05)

Recommendations

• Designing stable neural networks using convex analysis and ODEs
• Dynamical Systems–Based Neural Networks
• Probabilistic Lipschitz analysis of neural networks
• CLIP: cheap Lipschitz training of neural networks
• Regularisation of neural networks by enforcing Lipschitz continuity

Cites work

• Approximation of Lipschitz Functions Using Deep Spline Neural Networks
• scientific article; zbMATH DE number 432514 (Why is no real title available?)
• Learning the parts of objects by non-negative matrix factorization
• Lipschitz Certificates for Layered Network Structures Driven by Averaged Activation Operators
• Reluplex: an efficient SMT solver for verifying deep neural networks

Cited in

(1)

• Stability bounds for the unfolded forward-backward algorithm

This page was built for publication: ABBA neural networks: coping with positivity, expressivity, and robustness

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6583716)