Reinforcement Learning
Generalization
100

What an agent receives after performing an action 

Reward

100

One limitation of AlphaGo

Can't play anything besides Go/can't transfer skills from Go to a different game

200

The ability of a program to transfer what it has learned about one task to help it perform a different, related task

Transfer learning

200

What are Deep Q-learning systems still failing at/unable to achieve

abstraction/domain generalization/transfer learning

400

The game platform used by the DeepMind programmers in their work on reinforcement learning.

Atari 

M
e
n
u