Objectifs de la chaire
- Interactions entre algorithmes,
- Jeux hybrides entre des algorithmes et des agents rationnels,
- Apprentissage en ligne et problèmes de bandits,
- Réseaux de Neurones et GANs (Réseaux Antagonistes Génératifs),
- Décisions en environnements stratégiques complexes (problèmes de sélection compétitifs, bandits stratégiques à observation partielle, utilisation stratégique de l’information, congestion dans certains serveurs …)
- Optimisation globale Lipschitz à l’ordre 0.
Programme : IA Certifiable
Thème : optimisation et théorie des jeux pour l’IA
Porteur : Jérôme Renault, PR maths UT1-TSE
Co-chairs :
- Fabien Gensbittel, MCF maths UT1 -TSE
- Sébastien Gerchinovitz, IRT-DEEL
Porteur : Jérôme Renault, PR UT1 Capitole
Co-chairs :
- Fabien Gensbittel (UT1, TSE-R)
- Sebastien Gerchinovitz (UT3 IRT-Saint Exupery)
Doctorants :
- Maurizio d’Andrea (2020-2023)
- Etienne de Montbrun (2020-2023)
Post-docs :
- Tommaso Cesari (2019-2021)
- Dana Pizarro (2020-2022)
- Computer Science and data science department, University of Milan
- Department of Industrial Engineering- Universidad de Chile
- The Decision, Risk and Operations Division – Columbia University
- Business School- Universidad Torcuato Di Tella , Buenos Aires
- School of mathematical science, Tel-Aviv University
- Economic Dept, Toronto
- Economic Dept, University Roma Tor Vergata
- School of Management & Economics, University of Turin
Evénements organisés :
- ANITI Reinforcement Learning Virtual School (April-March 2021, 1500 participants de 43 pays)
- Workshop Machine Learning in Certified Systems, Theoretical and Practical Challenges (Toulouse et Montréal, Janvier 2021)
Autres temps forts :
- Co-edition des annales de l’International Society of Dynamic Games, Novembre 2020 Birkhauser.
-
Bilateral Trade: A Regret Minimization Perspective, Nicolò Cesa-Bianchi, Tommaso Cesari, Roberto Colomboni, Federico Fusco and Stefano Leonardi, Mathematics of Operations Research 2022, to appear.
-
Value-based distance between the information structures, F. Gensbittel, M. Peski and J. Renault. To appear in Theoretical Economics, 2022.
-
Long Information design, F. Koessler, M. Laclau, J. Renault and T. Tomala. Accepted in Theoretical Economics, 2022.
-
Splitting games over finite sets, F. Koessler, M. Laclau, J. Renault and T. Tomala. Minor revision, Maths Programming, 2022.
-
A New Theoretical Framework for Fast and Accurate Online Decision-Making. Nicolò Cesa- Bianchi, Tommaso Cesari, Yishay Mansour, Vianney Perchet. 2021. Thirty-fifth Conference on Neural Information Processing Systems (NeurIPS) 2021. https://hal.archives-ouvertes.fr/hal- 02976864
-
ROI Maximization in Stochastic Online Decision-Making, Nicolò Cesa-Bianchi, Tommaso Cesari, Yishay Mansour, Vianney Perchet. Advances in Neural Information Processing Systems 34 (NeurIPS), 2021.
-
The sample complexity of level set approximation François Bachoc, Tommaso Cesari, Sébastien Gerchinovitz Proceedings of the 24th International Conference on Artificial Intelligence and Statistics (AISTATS 2021, oral presentation), In press, 2021. https://hal.archives-ouvertes.fr/hal- 02976018
-
Numerical influence of ReLU’(0) on backpropagation David Bertoin, Jérôme Bolte, Sébastien Gerchinovitz et Edouard Pauwels, accepted at the Thirty-fifth Conference on Neural Information Processing Systems (NeurIPS 2021).
-
Instance-Dependent Bounds for Zeroth-order Lipschitz Optimization with Error Certificates. François Bachoc, Tommaso Cesari, Sébastien Gerchinovitz. 2021. Accepted at the Thirty- fifth Conference on Neural Information Processing Systems (NeurIPS 2021). https://hal.mines- ales.fr/ANITI/hal-03129721v3.
-
Investment Timing and Technological Breakthroughs, Jean-Paul Décamps, Fabien Gensbittel and Thomas Mariotti. TSE Working Paper, 21-1222, June 2021
-
Optimal Revenue Guarantees for Pricing in Large Markets”. Dana Pizarro, Jose Correa (Uni- versidad de Chile) and Victor Verdugo (Universidad de O’Higgins). Accepted to SAGT’21.
-
On a Competitive Selection Problem with Recall. Fabien Gensbittel, Dana Pizarro, Jérôme Renault. 31st European Conference on Operational Research (EURO), Athens July 2021
-
A Nearest Neighbor Characterization of Lebesgue Points in Metric Measure Spaces (pdf) T Cesari, R Colomboni 2021, Accepted at the Mathematical Statistics and Learning journal
-
An Efficient Algorithm for Cooperative Semi-Bandits (pdf) R Della Vecchia, T Cesari 2021, Accepted at The 32nd International Conference on Algorithmic Learning Theory (ALT 2021)
-
A Regret Analysis of Bilateral Trade. Nicolò Cesa-Bianchi, Tommaso Cesari, Roberto Colomboni, Federico Fusco, Stefano Leonardi. 2021. Accepted at The Twenty-Second ACM Conference on Economics and Computation (EC’21).
-
A Dynkin game on assets with incomplete information on the return, Tiziano De Angelis, Fabien Gensbittel, and Stéphane Villeneuve, Mathematics of Operations Research, vol. 46, (1), February 2021, p. 28-60.
-
Strategic Information transmission with sender’s approval, Jérôme Renault and Françoise Forges. International Journal of Game Theory, vol 50 (2), p. 475-502, June 2021.
-
Nonstochastic Bandits with Composite Anonymous Feedback, Nicolò Cesa-Bianchi, Tommaso Cesari, Roberto Colombani, Claudio Gentile, Yishay Mansour, submitted to JMLR 2021.
-
Cooperative online learning: Keeping your neighbors updated, Nicolò Cesa-Bianchi, Tommaso Cesari, Claire Monteleoni. ALT 2020 http://proceedings.mlr.press/v117/cesa-bianchi20a/cesa- bianchi20a.pdf
-
A High-Probability Safety Guarantee for Shifted Neural Network Surrogates. Mélanie Ducoffe, Sébastien Gerchinovitz, and Jayant Sen Gupta. Proceedings of SafeAI 2020, pages 74-82, 2020. http://ceur-ws.org/Vol-2560/paper11.pdf
-
Regret analysis of the Piyavskii-Shubert algorithm for global Lipschitz optimization. Clément Bouttier, Tommaso R. Cesari, and Sébastien Gerchinovitz. Preprint, 2020. https://arxiv.org/abs/2002.02390
-
Diversity-Preserving K-Armed Bandits, Revisited. Hédi Hadiji, Sébastien Gerchinovitz, Jean- Michel Loubes, and Gilles Stoltz. Preprint, 2020. https://hal.archives-ouvertes.fr/hal-02957485
-
Optimal Strategies in Zero-Sum Repeated Games with Incomplete Information: The Dependent Case, Fabien Gensbittel, and Miquel Oliu-Barton, Dynamic Games and Applications, vol. 10, February 2020, p. 819?835.
-
The value of observability in Dynamic Pricing, J. Correa, D. Pizarro and V. Verdugo. EC’20.
-
Limit Equilibrium Payoffs in Stochastic Games, J. Renault and B. Ziliotto. Mathematics of Operations Research 45(3), August 2020, p. 889-895.
-
Representation formulas for limit values of long run stochastic optimal controls, R. Buckdahn, J.Li , M. Quincampoix and J. Renault. SIAM journal of Control and Optimization 58 (4), p.1846-1873, 2020.
-
Acyclic gambling games, J. Renault and R. Laraki. Mathematics of Operations Research 45(4), November 2020, p.1237-1257.
-
Hidden Stochastic Games and Limit Equilibrium Payoffs, J. Renault and B. Ziliotto. Games and Economic Behavior, (124), p.122-139, November 2020.
-
Annals of the International Society of Dynamic Games Vol 17: Games of Conflict, Evolutionary Games, Economic Games, and Games Involving Common Interest, Novembre 2020 Birkhauser. Co-edition J. Renault and D. Ramsey. ISBN 978-3-030-56533-6. 346 pages
-
Book Mathematical Foundations of Game Theory, R. Laraki, J. Renault and S. Sorin. Universi- text book series, 229 p., Springer 2019.
En savoir +
La vie est un jeu, avec Jérôme Renault.
Comment la théorie des jeux devient-elle un outil de réflexion pour l’intelligence artificielle ? C’est le défi auquel s’attelle le mathématicien Jérôme Renault. L’occasion de découvrir de nouvelles façons de jouer. Interview.
French 
English