About Me

I'm an assistant professor at the Department of Electronics, Information and Bioengineering (DEIB) at Politecnico di Milano. I got my Ph.D. degree in Computer Science at Politecnico di Milano, where I was advised by Prof. Nicola Gatti. My research focuses on AI and Machine Learning, with particular emphasis on algorithmic game theory, which combines AI techniques with economic paradigms to build artificial agents capable of strategic reasoning in conflicting interactions. I am also interested in algorithms design, computational complexity, online learning and optimization.

You can find more details in my CV.

email  alberto DOT marchesi AT polimi DOT it

News

• chevron_right

  11.08.2023

  I have been awarded a PRIN 2022 grant and I am looking to hire a postdoc starting from around December 2023. Click here for more information!

• chevron_right

  21.06.2023

  2 Papers accepted at ICML 2023 + 1 Paper accepted at EC 2023!

• chevron_right

  03.11.2022

  Our paper "Simple Uncoupled No-Regret Learning Dynamics for Extensive-Form Correlated Equilibrium" (with Andrea Celli, Gabriele Farina and Nicola Gatti) has been accepted at the Journal of the ACM!

• chevron_right

  28.09.2022

  2 Papers accepted at NeurIPS 2022!

Publications

Conference Papers

• search Learning Optimal Contracts: How to Exploit Small Action Spaces

  Martino Bernasconi, Alberto Marchesi, Francesco Trovò
  The 27th International Conference on Artificial Intelligence and Statistics, AISTATS 2024, Valencia, Spain
  [Extended version]

• search Learning Optimal Contracts: How to Exploit Small Action Spaces

  Francesco Bacchiocchi, Matteo Castiglioni, Alberto Marchesi, Nicola Gatti
  The 12th International Conference on Learning Representations, ICLR 2024, Vienna, Austria
  [Extended version]

• search Finding Effective Ad Allocations: How to Exploit User History

  Matteo Castiglioni, Alberto Latino, Alberto Marchesi, Giulia Romano, Nicola Gatti, Chokha Palayamkottai
  The 23rd International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2024, Auckland, New Zealand
  [Extended version]

• search Persuading Farsighted Receivers in MDPs: the Power of Honesty

  Martino Bernasconi, Matteo Castiglioni, Alberto Marchesi, Mirco Mutti
  The 37th Conference on Neural Information Processing Systems, NeurIPS 2023, New Orleans, USA

• search Multi-Agent Contract Design: How to Commission Multiple Agents with Individual Outcomes

  Matteo Castiglioni, Alberto Marchesi, Nicola Gatti
  The 24th ACM Conference on Economics and Computation, EC 2023, London, UK

• search Optimal Rates and Efficient Algorithms for Online Bayesian Persuasion

  Martino Bernasconi, Matteo Castiglioni, Andrea Celli, Alberto Marchesi, Francesco Trovò, Nicola Gatti
  The 40th International Conference on Machine Learning, ICML 2023, Honolulu, USA
  

• search Constrained Phi-Equilibria

  Martino Bernasconi, Matteo Castiglioni, Alberto Marchesi, Francesco Trovò, Nicola Gatti
  The 40th International Conference on Machine Learning, ICML 2023, Honolulu, USA
  

• search A Unifying Framework for Online Optimization with Long-Term Constraints

  Matteo Castiglioni, Andrea Celli, Alberto Marchesi, Giulia Romano, Nicola Gatti
  The 36th Conference on Neural Information Processing Systems, NeurIPS 2022, New Orleans, USA
  [Supplemental Material] [Extended version]

• search Sequential Information Design: Learning to Persuade in the Dark

  Martino Bernasconi, Matteo Castiglioni, Alberto Marchesi, Nicola Gatti, Francesco Trovò
  The 36th Conference on Neural Information Processing Systems, NeurIPS 2022, New Orleans, USA
  [Supplemental Material] [Extended version]

• search Safe Learning in Tree-Form Sequential Decision Making: Handling Hard and Soft Constraints

  Martino Bernasconi, Federico Cacciamani, Matteo Castiglioni, Alberto Marchesi, Nicola Gatti, Francesco Trovò
  The 39th International Conference on Machine Learning, ICML 2022, Baltimore, USA
  

• search Designing Menus of Contracts Efficiently: The Power of Randomization

  Matteo Castiglioni, Alberto Marchesi, Nicola Gatti
  The 23rd ACM Conference on Economics and Computation, EC 2022, Boulder, USA
  [Extended version]

• search Public Signaling in Bayesian Ad Auctions

  Francesco Bacchiocchi, Matteo Castiglioni, Alberto Marchesi, Giulia Romano, Nicola Gatti
  The 31st International Joint Conference on Artificial Intelligence, IJCAI 2022, Vienna, Austria
  [Extended version]

• search The Power of Media Agencies in Ad Auctions: Improving Utility through Coordinated Bidding

  Giulia Romano, Matteo Castiglioni, Alberto Marchesi, Nicola Gatti
  The 31st International Joint Conference on Artificial Intelligence, IJCAI 2022, Vienna, Austria
  [Extended version]

• search Bayesian Persuasion Meets Mechanism Design: Going Beyond Intractability with Type Reporting

  Matteo Castiglioni, Alberto Marchesi, Nicola Gatti
  The 21st International Conference on Autonomous Agents and Multi-Agent Systems, AAMAS 2022, Virtual conference, Worldwide
  [Extended version]

• search Signaling in Posted Price Auctions

  Matteo Castiglioni, Giulia Romano, Alberto Marchesi, Nicola Gatti
  The 36th AAAI Conference on Artificial Intelligence, AAAI 2022, Virtual conference, Worldwide
  [Extended version]

• search Efficiency of Ad Auctions with Price Displaying

  Matteo Castiglioni, Diodato Ferraioli, Nicola Gatti, Alberto Marchesi, Giulia Romano
  The 36th AAAI Conference on Artificial Intelligence, AAAI 2022, Virtual conference, Worldwide
  [Extended version]

• search Exploiting Opponents Under Utility Constraints in Sequential Games

  Martino Bernasconi de Luca, Federico Cacciamani, Simone Fioravanti, Nicola Gatti, Alberto Marchesi, Francesco Trovò
  The 35th Conference on Neural Information Processing Systems, NeurIPS 2021, Virtual Conference, Worldwide
  [Supplemental Material]

• search Bayesian Agency: Linear versus Tractable Contracts

  Matteo Castiglioni, Alberto Marchesi, Nicola Gatti
  The 22nd ACM Conference on Economics and Computation, EC 2021, Virtual conference, Worldwide
  [Extended version]

• search Decentralized No-regret Learning Algorithms for Extensive-form Correlated Equilibria (Extended Abstract)

  Andrea Celli, Alberto Marchesi, Gabriele Farina, Nicola Gatti
  The 30th International Joint Conference on Artificial Intelligence, IJCAI 2021, Virtual Conference, Worldwide
  [Extended version]

• search Multi-Receiver Online Bayesian Persuasion

  Castiglioni Matteo, Marchesi Alberto, Celli Andrea, Gatti Nicola
  The 38th International Conference on Machine Learning, ICML 2021, Virtual conference, Worldwide
  [Supplemental Material]

• search Online Posted Pricing with Unknown Time-Discounted Valuations

  Giulia Romano, Gianluca Tartaglia, Alberto Marchesi, Nicola Gatti
  The 35th AAAI Conference on Artificial Intelligence, AAAI 2021, Virtual conference, Worldwide
  [Extended version]

• search Trembling-Hand Perfection and Correlation in Sequential Games

  Alberto Marchesi, Nicola Gatti
  The 35th AAAI Conference on Artificial Intelligence, AAAI 2021, Virtual conference, Worldwide
  [Extended version]

• search Signaling in Bayesian Network Congestion Games: the Subtle Power of Symmetry

  Matteo Castiglioni, Andrea Celli, Alberto Marchesi, Nicola Gatti
  The 35th AAAI Conference on Artificial Intelligence, AAAI 2021, Virtual conference, Worldwide

• search No-Regret Learning Dynamics for Extensive-Form Correlated Equilibrium

  Andrea Celli, Alberto Marchesi, Gabriele Farina, Nicola Gatti
  The 34th Conference on Neural Information Processing Systems, NeurIPS 2020, Virtual Conference, Worldwide
  whatshotwhatshotwhatshot  NeurIPS 2020 Best Paper Award, top 3 out of 9476 submissions
  [Supplemental Material]

• search Online Bayesian Persuasion

  Matteo Castiglioni, Andrea Celli, Alberto Marchesi, Nicola Gatti
  The 34th Conference on Neural Information Processing Systems, NeurIPS 2020, Virtual Conference, Worldwide
  Spotlight presentation, top 2.96% of submissions
  [Supplemental Material]

• search Learning Probably Approximately Correct Maximin Strategies in Simulation-Based Games with Infinite Strategy Spaces

  Alberto Marchesi, Francesco Trovò, Nicola Gatti
  The 19th International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2020, Virtual Conference, Worldwide
  [Extended version]

• search Learning to Correlate in Multi-Player General-Sum Sequential Games

  Andrea Celli, Alberto Marchesi, Tommaso Bianchi, Nicola Gatti
  The 33rd Conference on Neural Information Processing Systems, NeurIPS 2019, Vancouver, Canada
  [Supplemental Material]

• search Be a Leader or Become a Follower: The Strategy to Commit to with Multiple Leaders

  Matteo Castiglioni, Alberto Marchesi, Nicola Gatti
  The 28th International Joint Conference on Artificial Intelligence, IJCAI 2019, Macao, China
  [Extended version]

• search Leadership in Congestion Games: Multiple User Classes and Non-Singleton Actions

  Alberto Marchesi, Matteo Castiglioni, Nicola Gatti
  The 28th International Joint Conference on Artificial Intelligence, IJCAI 2019, Macao, China
  [Extended version]

• search Quasi-Perfect Stackelberg Equilibrium

  Alberto Marchesi, Gabriele Farina, Christian Kroer, Nicola Gatti, Tuomas Sandholm
  The 33rd AAAI Conference on Artificial Intelligence, AAAI 2019, Honolulu, USA
  [Extended version]

• search Trembling-Hand Perfection in Extensive-Form Games with Commitment

  Gabriele Farina, Alberto Marchesi, Christian Kroer, Nicola Gatti, Tuomas Sandholm
  The 27th International Joint Conference on Artificial Intelligence, IJCAI 2018, Stockholm, Sweden

• search Leadership in Singleton Congestion Games

  Alberto Marchesi, Stefano Coniglio, Nicola Gatti
  The 27th International Joint Conference on Artificial Intelligence, IJCAI 2018, Stockholm, Sweden
  

• search Computing the Strategy to Commit to in Polymatrix Games

  Giuseppe De Nittis, Alberto Marchesi, Nicola Gatti
  The 32nd AAAI Conference on Artificial Intelligence, AAAI 2018, New Orleans, USA
  [Extended version]

• search Pessimistic Leader-Follower Equilibria with Multiple Followers

  Stefano Coniglio, Nicola Gatti, Alberto Marchesi
  The 26th International Joint Conference on Artificial Intelligence, IJCAI 2017, Melbourne, Australia
  

• search On the Complexity of Nash Equilibrium Reoptimization

  Andrea Celli, Alberto Marchesi, Nicola Gatti
  The 33rd Conference on Uncertainty in Artificial Intelligence, UAI 2017, Sydney, Australia
  

• search Bilevel Programming Approaches to the Computation of Optimistic and Pessimistic Single-Leader-Multi-Follower Equilibria

  Nicola Basilico, Stefano Coniglio, Nicola Gatti, Alberto Marchesi
  The 16th International Symposium on Experimental Algorithms, SEA 2017, London, UK
  

Journal Papers

• search Designing menus of contracts efficiently: The power of randomization

  Matteo Castiglioni, Alberto Marchesi, Nicola Gatti
  Artificial Intelligence, 2023
  

• search Regret minimization in online Bayesian persuasion: Handling adversarial receiver's types under full and partial feedback models

  Matteo Castiglioni, Andrea Celli, Alberto Marchesi, Nicola Gatti
  Artificial Intelligence, 2023
  

• search Simple Uncoupled No-regret Learning Dynamics for Extensive-form Correlated Equilibrium

  Gabriele Farina, Andrea Celli, Alberto Marchesi, Nicola Gatti
  Journal of the ACM, Volume 69, 2022
  

• search Bayesian Agency: Linear versus Tractable Contracts

  Matteo Castiglioni, Alberto Marchesi, Nicola Gatti
  Artificial Intelligence, 2022
  

• search Committing to correlated strategies with multiple leaders

  Matteo Castiglioni, Alberto Marchesi, Nicola Gatti
  Artificial Intelligence, 2021
  

• search A characterization of quasi-perfect equilibria

  Nicola Gatti, Mario Gilli, Alberto Marchesi
  Games and Economic Behavior, 2020
  

• search Computing a Pessimistic Stackelberg Equilibrium with Multiple Followers: The Mixed-Pure Case

  Stefano Coniglio, Nicola Gatti, Alberto Marchesi
  Algorithmica, 2020
  

• search Bilevel programming methods for computing single-leader-multi-follower equilibria in normal-form and polymatrix games

  Nicola Basilico, Stefano Coniglio, Nicola Gatti, Alberto Marchesi
  EURO Journal on Computational Optimization, 2020
  

• search Leadership in singleton congestion games: What is hard and what is easy

  Matteo Castiglioni, Alberto Marchesi, Nicola Gatti, Stefano Coniglio
  Artificial Intelligence, 2019
  

Teaching

• link Informatica B

  Laurea Triennale in Ingegneria Meccanica e Laurea Triennale in Ingegneria Energetica (sez. SAM-ZZZ)
  Professor during the academic years 2022-2023; 2023-2024
  

• link Multi-agent Learning: From Theory to Practice

  Ph.D. in Information Technology and M.Sc. in Computer Science Engineering
  Professor during the academic year 2022-2023
  

• link Algorithmic Game Theory

  M.Sc. in Computer Science Engineering and M.Sc. in Mathematical Engineering
  TA during the academic years 2021-2022; 2022-2023; 2023-2024
  

• link Informatica A

  Laurea Triennale in Ingegneria Matematica (sez. M-Z)
  TA during the academic years 2018-2019; 2019-2020; 2020-2021; 2021-2022
  

• link Economics and Computation

  M.Sc. in Computer Science Engineering and M.Sc. in Mathematical Engineering
  TA during the academic years 2017-2018; 2018-2019; 2019-2020; 2020-2021
  

• link Game Theory

  M.Sc. in Mathematical Engineering
  TA during the academic year 2019-2020
  

Jobs

• person Postdoctoral Research Fellow

  I am looking for a postdoctoral research fellow to join my team at Politecnico di Milano and contribute to research in online learning and computational game theory. This is a fully funded, full-time research opportunity that does not involve any mandatory teaching duties. The contract duration for this position is expected to be 12 months. I offer a competitive salary, and applicants from outside Italy may have the opportunity to benefit from a researcher taxation scheme. This position is made possible through a collaborative research grant shared with Bocconi University. The successful candidate will be based at Politecnico di Milano working under my supervision, and they will also have the opportunity to closely collaborate with Andrea Celli's group at Bocconi University. The research project considers topics at the intersection of online learning and computational game theory. The primary objective of the project is to delve into the challenges of steering the collective behaviour of no-regret learners towards specific target outcomes. The project is a natural continuation of some of my recent research. Ideal candidates should have a track record in performing top-quality research in machine learning theory/computational game theory compatible with the standards of venues like EC, ICML, and NeurIPS. Demonstrated past experience with such topics will be evaluated very positively.
  If you are interested in this postdoctoral position and seek further information, please do not hesitate to reach out to me at alberto.marchesi@polimi.it. Please attach an up-to-date CV including a list of publications.