Sign in

Q-Ensemble for Offline RL: Don't Scale the Ensemble, Scale the Batch Size.

Alexander NikulinVladislav KurenkovDenis TarasovDmitry AkimovSergey Kolesnikov
Published in: CoRR (2022)
Keyphrases
  • learning algorithm
  • reinforcement learning
  • training data
  • learning process
  • dynamic programming
  • batch size