Paper
Neural Episodic Control
Deep reinforcement learning methods attain super-human performance in a wide range of environments. Such methods are grossly inefficient, often taking orders of magnitudes more data than humans to achieve reasonable performance. We propose Neural Episodic Control: a deep reinforcement learning agent that is able to rapidly assimilate new experiences and act upon them. Our agent uses a semi-tabular representation of the value function: a buffer of past experience containing slowly changing state representations and rapidly updated estimates of the value function. We show across a wide range of environments that our agent learns significantly faster than other state-of-the-art, general purpose deep reinforcement learning agents.
Authors: Pritzel, Alexander · Uria, Benigno · Srinivasan, Sriram · Puigdomènech, Adrià · Vinyals, Oriol · Hassabis, Demis · Wierstra, Daan · Blundell, Charles