Markov decision processes

Needs first

• Probability and statistics

Helps first

Nothing recommended beyond what is required.

Unlocks

• Value-based methods
• Policy gradient methods