Playing Atari Ball Games with Hierarchical Reinforcement Learning

Hua Huang, Adrian Barbu

Human beings are particularly good at reasoning and inference from just a few examples. When facing new tasks, humans will leverage knowledge and skills learned before, and quickly…