no code implementations • 3 Oct 2021 • Chi Zhang, Sanmukh Rao Kuppannagari, Viktor K Prasanna
Current implementations exhibit poor performance due to challenges such as irregular memory accesses and thread-level synchronization overheads on CPU.
1 code implementation • 2 Oct 2021 • Chi Zhang, Sanmukh Rao Kuppannagari, Viktor K Prasanna
This leads to large overestimations of the Q values and performance deterioration of the learned policy.
no code implementations • 1 Jan 2021 • Chi Zhang, Sanmukh Rao Kuppannagari, Viktor Prasanna
The goal of Offline Reinforcement Learning (RL) is to address this problem by learning effective policies using previously collected datasets.
no code implementations • 8 Jun 2020 • Chi Zhang, Sanmukh Rao Kuppannagari, Viktor K. Prasanna
Furthermore, we propose to generate \emph{diverse} model rollouts by non-uniform sampling of the environment states such that the entropy of the model rollouts is maximized.
Model-based Reinforcement Learning reinforcement-learning +1