• Identifying Reusable Early-Life Options 

      Weber, Aline (2020) [Trabalho de conclusão de graduação]
      We introduce a method for identifying short-duration reusable motor behaviors, which we call early-life options, that allow robots to perform well even in the very early stages of their lives. This is important when agents ...
    • O-MuZero : abstract planning models Induced by Options on the MuZero Algorithm 

      Jacobi, Otavio Flores (2021) [Trabalho de conclusão de graduação]
      Training Reinforcement Learning agents that learn both the value function and the envi ronment model can be a very time consuming method, one of the main reasons for that is that these agents learn by actions one step at ...