Research

Tutte le verità sono facili da capire una volta che sono state rivelate. Il difficile è scoprile. (All truths are easy to understand once they are discovered. The point is to discover them.)
Galileo Galilei, 1564-1642


PhD Thesis

Exploration–exploitation dilemma in Reinforcement Learning under various form of prior knowledge, November 6th 2019, Inria Lille, France [PDF]


Conference Papers

  • Ronald Ortner, Matteo Pirotta, Alessandro Lazaric, Ronan Fruit, Odalric Maillard
    Regret Bounds for Learning State Representations in Reinforcement Learning
    NeurIPS 2019, Vancouver, Canada. [PDF]
    Note: Unfortunately, the proofs of the regret bounds for the schemes given in Sections 5 and 6.2 (when guessing the diameter resp. the size of the state space) have a gap that seems not easy to close. The other results are however not affected.
  • Jian Qian, Ronan Fruit, Matteo Pirotta, Alessandro Lazaric
    Exploration Bonus for Regret Minimization in Discrete and Continuous Average Markov Decision Processes
    NeurIPS 2019, Vancouver, Canada. [PDF, poster]
  • Ronan Fruit, Matteo Pirotta, Alessandro Lazaric:
    Near Optimal Exploration-Exploitation in Non-Communicating Markov Decision Processes.
    NeurIPS 2018, Montreal, Canada. [PDF, Poster, Presentation (00:20:45)]
  • Ronan Fruit, Matteo Pirotta, Alessandro Lazaric,Ronald Ortner:
    Efficient Bias-Span-Constrained Exploration-Exploitation in Reinforcement Learning.
    ICML 2018, Stockholm, Sweden. [PDF, Poster, Presentation (01:03:00)]
  • Ronan Fruit, Matteo Pirotta, Alessandro Lazaric, Emma Brunskill:
    Regret Minimization in MDPs with Options without Prior Knowledge.
    NeurIPS 2017, Long Beach, California, USA. [PDF, Poster, Presentation]
  • Ronan Fruit, Alessandro Lazaric:
    Exploration–Exploitation in MDPs with Options.
    AISTATS 2017, Fort Lauderdale, Florida, USA. [PDF, Poster]

Workshop Papers


Visit and Talks


Program Committees