Source/tutorial: https://lkieliger.medium.com/deep-reinforcement-learning-in-practice-by-playing-doom-part-1-getting-started-618c99075c77
ViZDoom: https://github.com/Farama-Foundation/ViZDoom
train: https://www.youtube.com/watch?v=jMPmPrutask
test: https://www.youtube.com/watch?v=n4QLKQDIuHg
Reinforcement Learning has 4 key elements: agent, reward, environment, and action. In this project, doom guy (player) is the agent as he can take some action (for example: shooting, moving left, moving right) inside of the game environment then depending on the result of doom guy’s action he might get a reward. The AI that’s controlling doom guy learns what actions to take inside of the environment to maximize reward. In this case, kill the enemy and minimize steps taken to do so.