Provably Efficient Algorithms for Large-Scale Reinforcement Learning

Informazioni relative al progetto

SCALER

ID dell’accordo di sovvenzione: 950180

DOI

10.3030/950180

Data della firma CE 3 Settembre 2020

Data di avvio 1 Ottobre 2021

Data di completamento 30 Settembre 2026

Finanziato da

• EXCELLENT SCIENCE - European Research Council (ERC)

Costo totale

€ 1 493 990,00

Contributo UE

€ 1 493 990,00

Coordinato da

UNIVERSIDAD POMPEU FABRA
Spain

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 .

Risultati finali

Open Research Data Pilot (1)

Data Management Plan (DMP)

The Open Research Data Pilot will be prepared and submitted to the European Commission

Pubblicazioni

Conference proceedings (14)

Offline Primal-Dual Reinforcement Learning for Linear MDPs

Autori: G. Gabbianelli, G. Neu, N. Okolo, M. Papini
Pubblicato in: Proceedings of the Twenty-seventh International Conference on Artificial Intelligence and Statistics (AISTATS), 2024
Editore: Proceedings of Machine Learning Research

Scalable Representation Learning in Linear Contextual Bandits with Constant Regret Guarantees

Autori: Tirinzoni A.; Papini M.; Touati A.; Lazaric A.; Pirotta M.
Pubblicato in: Advances in Neural Information Processing Systems 35 (NeurIPS 2022), 2022
Editore: NeurIPS foundation
DOI: 10.48550/arxiv.2210.13083

Efficient Global Planning in Large MDPs via Stochastic Primal-Dual Optimization

Autori: Gergely Neu, Nneka Okolo
Pubblicato in: Proceedings of The 34th International Conference on Algorithmic Learning Theory (ALT 2023), 2023
Editore: Proceedings of Machine Learning Research

Lifting the Information Ratio: An Information-Theoretic Analysis of Thompson Sampling for Contextual Bandits

Autori: Gergely Neu, Julia Olkhovskaya, Matteo Papini, Ludovic Schwartz
Pubblicato in: Advances in Neural Information Processing Systems 35 (NeurIPS 2022), 2022
Editore: NeurIPS foundation

Nonstochastic Contextual Combinatorial Bandits

Autori: L. Zierahn, D. van der Hoeven, N. Cesa-Bianchi, G. Neu
Pubblicato in: Proceedings of the Twenty-sixth International Conference on Artificial Intelligence and Statistics (AISTATS), 2023
Editore: Proceedings of Machine Learning Research

Optimistic Information-Directed Sampling

Autori: G. Neu, M. Papini, L. Schwartz
Pubblicato in: Proceedings of the 36th Annual Conference on Learning Theory (COLT), 2024
Editore: Proceedings of Machine Learning Research

Dealing with Unbounded Gradients in Stochastic Saddle-Point Optimizaiton

Autori: G. Neu, N. Okolo
Pubblicato in: Proceedings of the 41st International Conference on Machine Learning (ICML), 2024
Editore: Proceedings of Machine Learning Research

Proximal Point Imitation Learning

Autori: Luca Viano, Angeliki Kamoutsi, Gergely Neu, Igor Krawczuk, Volkan Cevher
Pubblicato in: Advances in Neural Information Processing Systems 35 (NeurIPS 2022), 2022
Editore: NeurIPS foundation

Importance-Weighted Offline Learning Done Right

Autori: G. Gabbianelli, G. Neu, M. Papini
Pubblicato in: Proceedings of the 34th International Conference on Algorithmic Learning Theory (ALT), 2024
Editore: Proceedings of Machine Learning Research

Online learning with off-policy feedback

Autori: Germano Gabbianelli, Matteo Papini, Gergely Neu
Pubblicato in: Proceedings of The 34th International Conference on Algorithmic Learning Theory (ALT 2023), 2023
Editore: Proceedings of Machine Learning Research

First-and Second-Order Bounds for Adversarial Linear Contextual Bandits

Autori: J. Olkhovskaya, J. Mayo, T. van Erven, G. Neu, C.-Y. Wei
Pubblicato in: Advances in Neural Information Processing Systems 36 (NeurIPS), 2023
Editore: NeurIPS foundation

Optimistic Planning by Regularized Dynamic Programming

Autori: Antoine Moulin, Gergely Neu
Pubblicato in: International Conference on Machine Learning (ICML 2022), 2023
Editore: Proceedings of Machine Learning Research

Generalization bounds via convex analysis

Autori: Gabor Lugosi, Gergely Neu
Pubblicato in: Proceedings of Thirty Fifth Conference on Learning Theory (COLT 2022), 2022
Editore: Proceedings of Machine Learning Research

Adversarial Contextual Bandits Go Kernelized

Autori: G. Neu, J. Olkhovskaya, S. Vakili
Pubblicato in: Proceedings of the 34th International Conference on Algorithmic Learning Theory (ALT), 2024
Editore: Proceedings of Machine Learning Research

Peer reviewed articles (2)

Smoothing policies and safe policy gradients

Autori: Matteo Papini; Matteo Pirotta; Marcello Restelli
Pubblicato in: Machine Learning, Numero 111, 2022, Pagina/e 4081–4137, ISSN 1573-0565
Editore: Springer
DOI: 10.1007/s10994-022-06232-6

A note on regularised NTK dynamics with an application to PAC-Bayesian training

Autori: Clerico, Eugenio; Guedj, Benjamin
Pubblicato in: Transactions on Machine Learning Research, 2024, ISSN 2835-8856
Editore: Transactions on Machine Learning Research
DOI: 10.48550/arxiv.2312.13259