Control for Deep and Federated Learning

Informazioni relative al progetto

CoDeFeL

ID dell’accordo di sovvenzione: 101096251

DOI

10.3030/101096251

Data della firma CE 10 Ottobre 2023

Data di avvio 1 Settembre 2024

Data di completamento 31 Agosto 2029

Finanziato da

• European Research Council (ERC)

Costo totale

€ 2 499 224,00

Contributo UE

€ 2 499 224,00

Coordinato da

FRIEDRICH-ALEXANDER-UNIVERSITAET ERLANGEN-NUERNBERG
Germany

CORDIS fornisce collegamenti ai risultati finali pubblici e alle pubblicazioni dei progetti ORIZZONTE.

I link ai risultati e alle pubblicazioni dei progetti del 7° PQ, così come i link ad alcuni tipi di risultati specifici come dataset e software, sono recuperati dinamicamente da .OpenAIRE .

Pubblicazioni

Conference proceedings (2)

Machine Learning and Control: Foundations, Advances, and Perspectives

Autori: Enrique Zuazua
Pubblicato in: Proceedings of the International Congress of Mathematicians 2026 - Volume 7: Invited Lectures (Sections 15–20), 2026
Editore: Society for Industrial and Applied Mathematics
DOI: 10.48550/ARXIV.2510.03303

Shape and Topology Optimization for Network of Timoshenko Beams

Autori: Meizhi Qian, Jan Sokołowski
Pubblicato in: 2025 29th International Conference on Methods and Models in Automation and Robotics (MMAR), 2025
Editore: IEEE
DOI: 10.1109/MMAR65820.2025.11150880

Peer reviewed articles (24)

Moments, time-inversion and source identification for the heat equation

Autori: Kang Liu, Enrique Zuazua
Pubblicato in: Inverse Problems, Numero 42, 2026, ISSN 0266-5611
Editore: IOP Publishing
DOI: 10.1088/1361-6420/AE3797

Representation and regression problems in neural networks: Relaxation, generalization, and numerics

Autori: Kang Liu, Enrique Zuazua
Pubblicato in: Mathematical Models and Methods in Applied Sciences, Numero 35, 2025, ISSN 0218-2025
Editore: World Scientific Pub Co Pte Ltd
DOI: 10.1142/S0218202525500228

Nonlinear equilibrium transitions in a potential game model for federated learning

Autori: Kang Liu, Ziqi Wang, Enrique Zuazua
Pubblicato in: Physica D: Nonlinear Phenomena, Numero 495, 2026, ISSN 0167-2789
Editore: Elsevier BV
DOI: 10.1016/J.PHYSD.2026.135288

On control networks over finite lattices

Autori: Zhengping Ji, Daizhan Cheng
Pubblicato in: Science China Information Sciences, Numero 68, 2025, ISSN 1674-733X
Editore: Springer Science and Business Media LLC
DOI: 10.1007/S11432-024-4252-7

Universal Approximation of Dynamical Systems by Semiautonomous Neural ODEs and Applications

Autori: Ziqian Li, Kang Liu, Lorenzo Liverani, Enrique Zuazua
Pubblicato in: SIAM Journal on Numerical Analysis, Numero 64, 2026, ISSN 0036-1429
Editore: Society for Industrial & Applied Mathematics (SIAM)
DOI: 10.1137/24M1679690

Approximate and Weighted Data Reconstruction Attack in Federated Learning

Autori: Yongcun Song, Ziqi Wang, Enrique Zuazua
Pubblicato in: IEEE Transactions on Big Data, Numero 12, 2026, ISSN 2332-7790
Editore: Institute of Electrical and Electronics Engineers (IEEE)
DOI: 10.48550/ARXIV.2308.06822

Random batch methods for discretized PDEs on graphs

Autori: Martin Hernandez, Enrique Zuazua
Pubblicato in: Journal of Computational Physics, Numero 569, 2026, ISSN 0021-9991
Editore: Elsevier BV
DOI: 10.48550/ARXIV.2506.11809

Tracking Controllability for Finite-Dimensional Linear Systems

Autori: Sebastian Zamorano, Enrique Zuazua
Pubblicato in: IEEE Transactions on Automatic Control, 2026, ISSN 0018-9286
Editore: Institute of Electrical and Electronics Engineers (IEEE)
DOI: 10.48550/ARXIV.2407.18641

Large-time asymptotics for hyperbolic systems with non-symmetric relaxation: An algorithmic approach

Autori: Timothée Crin-Barat, Lorenzo Liverani, Ling-Yun Shou, Enrique Zuazua
Pubblicato in: Journal de Mathématiques Pures et Appliquées, Numero 202, 2025, ISSN 0021-7824
Editore: Elsevier BV
DOI: 10.1016/J.MATPUR.2025.103757

Regional and partial observability and control of waves

Autori: Belhassen Dehman, Sylvain Ervedoza, Enrique Zuazua
Pubblicato in: Comptes Rendus. Mathématique, Numero 363, 2025, ISSN 1631-073X
Editore: MathDoc/Centre Mersenne
DOI: 10.5802/CRMATH.805

Invariant and Dual-Invariant Subspaces of k-Valued Networks

Autori: Daizhan Cheng, Hongsheng Qi, Xiao Zhang, Zhengping Ji
Pubblicato in: Journal of Systems Science and Complexity, Numero 38, 2025, ISSN 1009-6124
Editore: Springer Science and Business Media LLC
DOI: 10.1007/S11424-025-4411-1

"Optimal <mml:math xmlns:mml=""http://www.w3.org/1998/Math/MathML"" altimg=""si17.svg"" display=""inline"" id=""d1e336""><mml:msup><mml:mrow><mml:mi>L</mml:mi></mml:mrow><mml:mrow><mml:mi>p</mml:mi></mml:mrow></mml:msup></mml:math>-approximation of convex sets by convex subsets"

Autori: Zakaria Fattah, Ilias Ftouhi, Enrique Zuazua
Pubblicato in: Nonlinear Analysis, Numero 261, 2025, ISSN 0362-546X
Editore: Elsevier BV
DOI: 10.1016/J.NA.2025.113866

Clustering in Pure-Attention Hardmax Transformers and Its Role in Sentiment Analysis

Autori: Albert Alcalde, Giovanni Fantuzzi, Enrique Zuazua
Pubblicato in: SIAM Journal on Mathematics of Data Science, Numero 7, 2026, ISSN 2577-0187
Editore: Society for Industrial & Applied Mathematics (SIAM)
DOI: 10.1137/24M167086X

Cluster-Based Classification with Neural ODEs via Control

Autori: Antonio Álvarez-Lόpez, Rafael Orive-Illera, Enrique Zuazua
Pubblicato in: Journal of Machine Learning, Numero 4, 2026, ISSN 2790-203X
Editore: Global Science Press
DOI: 10.4208/JML.241114

Observer-Based Realization of Control Systems

Autori: Daizhan Cheng, Xiao Zhang, Zhengping Ji, Changxi Li
Pubblicato in: IEEE Transactions on Automatic Control, Numero 71, 2026, ISSN 0018-9286
Editore: Institute of Electrical and Electronics Engineers (IEEE)
DOI: 10.1109/TAC.2025.3604656

Exact sequence interpolation with transformers

Autori: Albert Alcalde, Giovanni Fantuzzi, Enrique Zuazua
Pubblicato in: Mathematical Foundations of Machine Learning, Numero 2, 2026, ISSN 3059-3492
Editore: Springer Science and Business Media LLC
DOI: 10.48550/ARXIV.2502.02270

Pointwise constraints for scalar conservation laws with positive wave velocity

Autori: Nicola De Nitti, Denis Serre, Enrique Zuazua
Pubblicato in: Zeitschrift für angewandte Mathematik und Physik, Numero 76, 2025, ISSN 0044-2275
Editore: Springer Science and Business Media LLC
DOI: 10.1007/S00033-025-02459-0

Topological Derivative Method for Design and Control of Timoshenko Beam Networks

Autori: Meizhi Qian, Jairo Rocha de Faria, Antonio J. B. Santos, Jan Sokołowski, Ana P. P. Wyse
Pubblicato in: Applied Mathematics &amp; Optimization, Numero 93, 2026, ISSN 0095-4616
Editore: Springer Science and Business Media LLC
DOI: 10.1007/S00245-025-10369-W

Sharp numerical approximation of the Hardy constant

Autori: Liviu I. Ignat, Enrique Zuazua
Pubblicato in: Discrete and Continuous Dynamical Systems, Numero 54, 2026, ISSN 1078-0947
Editore: American Institute of Mathematical Sciences (AIMS)
DOI: 10.3934/DCDS.2025165

FedADMM-InSa: An inexact and self-adaptive ADMM for federated learning

Autori: Yongcun Song, Ziqi Wang, Enrique Zuazua
Pubblicato in: Neural Networks, Numero 181, 2025, ISSN 0893-6080
Editore: Elsevier BV
DOI: 10.1016/J.NEUNET.2024.106772

Boundary sidewise observability of the wave equation

Autori: Belhassen Dehman, Enrique Zuazua
Pubblicato in: Journal of the European Mathematical Society, 2026, ISSN 1435-9855
Editore: European Mathematical Society - EMS - Publishing House GmbH
DOI: 10.4171/JEMS/1688

On universal eigenvalues and eigenvectors of hypermatrices

Autori: Daizhan Cheng, Zhengping Ji
Pubblicato in: Journal of the Franklin Institute, Numero 362, 2025, ISSN 0016-0032
Editore: Elsevier BV
DOI: 10.1016/J.JFRANKLIN.2025.108126

Deep neural ODE operator networks for PDEs

Autori: Ziqian Li, Kang Liu, Yongcun Song, Hangrui Yue, Enrique Zuazua
Pubblicato in: Mathematical Models and Methods in Applied Sciences, Numero 36, 2026, ISSN 0218-2025
Editore: World Scientific Pub Co Pte Ltd
DOI: 10.1142/S0218202526420054

Control of hyperbolic and parabolic equations on networks and singular limits

Autori: Jon Asier Bárcena-Petisco, Márcio Cavalcante, Giuseppe Maria Coclite, Nicola De Nitti, Enrique Zuazua
Pubblicato in: Mathematical Control and Related Fields, Numero 15, 2025, ISSN 2156-8472
Editore: American Institute of Mathematical Sciences (AIMS)
DOI: 10.3934/MCRF.2024015