A testbed of Atari games for measuring how well AI agents learn.
ALE is an old-school arcade exam for AI. The AI only gets a screen and a joystick. Then it learns not to stink.
Researchers use it to test RL. The same Atari games keep the contest fair.
RL
ALE gives RL the same game setup for fair algorithm tests.
Deep Q-Network
Deep Q-Network used ALE to prove pixels can be enough for game learning.
OpenAI Gym
OpenAI Gym often wraps ALE in a standard way to run it.
Agent
An Agent watches the screen, presses buttons, and learns from the score.