Roberto Colomboni

Lecturer in Statistical Science and AI

School of Mathematics, University of Bristol

email: roberto [dot] <you-know-what-goes-here> [at] bristol [dot] ac [dot] uk

Curriculum vitae

Research interests

Online Learning, Multi-Armed Bandits, Statistical Learning, Optimisation, Digital Markets

In a nutshell...

I am a Lecturer in Statistical Science and AI at the University of Bristol.

My research is mostly theoretical, focusing on online learning, optimisation, and statistical learning theory: I study how to design learning algorithms and prove sharp performance guarantees that explain when and why they work.

I also have a sweet tooth for problems arising in digital markets, such as dynamic pricing, auctions, brokerage, and market making: I study the design of learning algorithms and market mechanisms that promote efficiency, welfare, and fairness in sequential decision-making problems involving uncertainty, limited feedback, and strategic behaviour, with applications to online marketplaces, mobility platforms, and digital advertising.

Preprints

• G. Beretta, T. Cesari, R. Colomboni, A. Paudice
  New Bounds for the Last Iterate of the Stochastic subGradient Method
  arXiv

• F. Bachoc, R. Colomboni, E. Kaufmann
  Repeated Bilateral Trade: The Quest for Fairness
  arXiv

• T. Cesari and R. Colomboni
  Two-Action Apple Tasting with Switching Costs
  arXiv

• T. Cesari and R. Colomboni
  Optimal Gap-Dependent Regret for Private Stochastic Decision-Theoretic Online Learning
  arXiv

• G. Beretta, T. Cesari, and R. Colomboni
  Generalized Pinsker inequality for Bregman divergences of negative Tsallis entropies
  arXiv

• F. Bachoc, N. Cesa-Bianchi, T. Cesari, and R. Colomboni
  Online budget allocation with censored semi-bandit feedback
  arXiv

Publications

See also Google Scholar.

• F. Bacchiocchi, M. Castiglioni, R. Colomboni, and A. Marchesi
  Online bilateral trade with minimal feedback: don't waste seller's time
  The Thirty-ninth Annual Conference on Neural Information Processing Systems (NeurIPS 2025), 2025.

• N. Bolić, T. Cesari, R. Colomboni, and C. Paravalos
  Online learning in the repeated mediated newsvendor problem
  The Thirty-ninth Annual Conference on Neural Information Processing Systems (NeurIPS 2025), 2025.

• N. Cesa-Bianchi, T. Cesari, R. Colomboni, L. Foscari, V. Pathak
  Market making without regret
  Proceedings of Thirty Eighth Conference on Learning Theory (COLT 2025), Volume 291, Pages 799--837, PMLR, 2025.
  (Cite: BibTeX here)

• F. Bachoc, T. Cesari, R. Colomboni
  A parametric contextual online learning theory of brokerage
  Forty-second International Conference on Machine Learning (ICML 2025), 2025.
  (Cite: BibTeX here)

• N. Cesa-Bianchi, R. Colomboni, and M. Kasy
  Adaptive maximization of social welfare
  Econometrica. Volume 93, Number 3, Pages 1073-1104. 2025
  (Cite: BibTeX here)

• T. Cesari and R. Colomboni
  An online learning theory of trading-volume maximization
  The Thirteenth International Conference on Learning Representations (ICLR 2025), 2025
  (Cite: BibTeX here)

• F. Bachoc, T. Cesari, and R. Colomboni
  A tight regret analysis of non-parametric repeated contextual brokerage
  The 28th International Conference on Artificial Intelligence and Statistics (AISTATS 2025), 2025.
  (Cite: BibTeX here)

• F. Bachoc, N. Cesa-Bianchi, T. Cesari, and R. Colomboni
  Fair online bilateral trade
  The Thirty-eighth Annual Conference on Neural Information Processing Systems (NeurIPS 2024), 2024.
  (Cite: BibTeX here)

• R. Colomboni, E. Esposito, and A. Paudice
  An improved uniform convergence bound with fat-shattering dimension
  Information Processing Letters.106539, ISSN 0020-0190, Elsevier, 2024.
  (Cite: BibTeX here)

• F. Bachoc, T. Cesari, R. Colomboni, and A. Paudice
  A theoretical framework for zeroth-order budget convex optimization
  Transactions on Machine Learning Research (TMLR). 2024.
  (Cite: BibTeX)

• N. Cesa-Bianchi, T. Cesari, R. Colomboni, F. Fusco, and S. Leonardi
  Regret analysis of bilateral trade with a smoothed adversary
  Journal of Machine Learning Research (JMLR). Volume 25, Number 234, Pages 1–36. 2024.
  (Cite: BibTeX)

• N. Cesa-Bianchi, T. Cesari, R. Colomboni, F. Fusco, and S. Leonardi
  The role of transparency in repeated first-price auctions with unknown valuations
  Proceedings of the 56th Annual ACM Symposium on Theory of Computing (STOC 2024), Pages 225-236, Association for Computing Machinery, 2024
  (Cite: BibTeX here)

• N. Bolić, T. Cesari, and R. Colomboni
  An online learning theory of brokerage
  Proceedings of the 23rd International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2024), Pages 216–224, 2024.
  (Cite: BibTeX here)

• N. Cesa-Bianchi, T. Cesari, R. Colomboni, F. Fusco, and S. Leonardi
  Repeated bilateral trade against a smoothed adversary
  Proceedings of Thirty Sixth Conference on Learning Theory (COLT 2023), Volume 195, Pages 1095-1130, PMLR, 2023.
  (Cite: BibTeX here)

• N. Cesa-Bianchi, T. Cesari, R. Colomboni, F. Fusco, and S. Leonardi
  Bilateral trade: a regret minimization perspective
  Mathematics of Operations Research, 2023.
  (Cite: BibTeX here)

• N. Cesa-Bianchi, T. Cesari, R. Colomboni, C. Gentile, and Y. Mansour
  Nonstochastic bandits with composite anonymous feedback
  Journal of Machine Learning Research (JMLR). Volume 23, Number 277, Pages 1–24. 2022.
  (Cite: BibTeX)

• N. Cesa-Bianchi, T. Cesari, R. Colomboni, F. Fusco, and S. Leonardi
  A regret analysis of bilateral trade
  The 22nd ACM Conference on Economics and Computation (EC 2021). Pages 289–309. ACM Press, 2021.
  (Cite: BibTeX)

• T. Cesari and R. Colomboni
  A nearest neighbor characterization of Lebesgue points in metric measure spaces
  Mathematical Statistics and Learning, Volume 3, Issue 1, Pages 71–112. EMS Press, 2021.
  (Cite: BibTeX)