Publications

Conference and journal papers

Papini, Matteo, Andrea Tirinzoni, Aldo Pacchiano, Marcello Restelli, Alessandro Lazaric, and Matteo Pirotta. 2021. “Reinforcement Learning in Linear MDPs: Constant Regret and Representation Selection.” To appear at NeurIPS 2021.

Papini, Matteo, Andrea Tirinzoni, Marcello Restelli, Alessandro Lazaric, and Matteo Pirotta. 2021. “Leveraging Good Representations in Linear Contextual Bandits.” In ICML, 139:8371–80. Proceedings of Machine Learning Research. PMLR.

Metelli, Alberto Maria, Matteo Papini, Pierluca D’Oro, and Marcello Restelli. 2021. “Policy Optimization as Online Learning with Mediator Feedback.” In AAAI, 8958–66. AAAI Press.

Papini, Matteo, Andrea Battistello, and Marcello Restelli. 2020. “Balancing Learning Speed and Stability in Policy Gradient via Adaptive Exploration.” In AISTATS, 108:1188–99. Proceedings of Machine Learning Research. PMLR.

Metelli, Alberto Maria, Matteo Papini, Nico Montali, and Marcello Restelli. 2020. “Importance Sampling Techniques for Policy Optimization.” J. Mach. Learn. Res. 21: 141:1–75.

D’Oro, Pierluca, Alberto Maria Metelli, Andrea Tirinzoni, Matteo Papini, and Marcello Restelli. 2020. “Gradient-Aware Model-Based Policy Search.” In AAAI, 3801–8. AAAI Press.

Bisi, Lorenzo, Luca Sabbioni, Edoardo Vittori, Matteo Papini, and Marcello Restelli. 2020. “Risk-Averse Trust Region Optimization for Reward-Volatility Reduction.” In IJCAI, 4583–89. ijcai.org.

Papini, Matteo, Alberto Maria Metelli, Lorenzo Lupo, and Marcello Restelli. 2019. “Optimistic Policy Optimization via Multiple Importance Sampling.” In ICML, 97:4989–99. Proceedings of Machine Learning Research. PMLR.

Beraha, Mario, Alberto Maria Metelli, Matteo Papini, Andrea Tirinzoni, and Marcello Restelli. 2019. “Feature Selection via Mutual Information: New Theoretical Insights.” In IJCNN, 1–9. IEEE.

Papini, Matteo, Damiano Binaghi, Giuseppe Canonaco, Matteo Pirotta, and Marcello Restelli. 2018. “Stochastic Variance-Reduced Policy Gradient.” In ICML, 80:4023–32. Proceedings of Machine Learning Research. PMLR.

Metelli, Alberto Maria, Matteo Papini, Francesco Faccio, and Marcello Restelli. 2018. “Policy Optimization via Importance Sampling.” In NeurIPS, 5447–59.

Papini, Matteo, Matteo Pirotta, and Marcello Restelli. 2017. “Adaptive Batch Size for Safe Policy Gradients.” In NeurIPS, 3591–3600.

Workshop papers

Alessandro Gianola, Marco Montali, and Matteo Papini. 2021. “Automated Reasoning for Reinforcement Learning Agents in Structured Environments.” OVERLAY workshop on fOrmal VERification, Logic, Automata and sYnthesis, Padova, Italy.

Matteo Papini, Andrea Tirinzoni, Aldo Pacchiano, Marcello Restelli, Alessandro Lazaric, and Matteo Pirotta. 2021. “Reinforcement Learning in Linear MDPs: Constant Regret and Representation Selection.” ICML Workshop on Reinforcement Learning Theory (virtual)

Matteo Papini, Andrea Battistello, and Marcello Restelli. 2019. “Safe Exploration in Gaussian Policy Gradient.” NeurIPS-2019 Workshop on Safety and Robustness in Decision Making, Vancouver, Canada.

Matteo Papini, Andrea Battistello, and Marcello Restelli. 2018. “Safely Exploring Policy Gradient.” 14th European Workshop on Reinforcement Learning, Lille, France.