Reinforcement Learning

An Introduction

second edition by Richard S. Sutton and Andrew G. Barto

issues
1. blackjack
2. section 6.2 batch td0
3. expected sarsa not obviously outperform sarsa

todo
1. figure 5.3
2. figure 5.4
3. exercise 5.12
4. exercise 5.14
5. off-policy mc prediction
6. off-policy mc control
7. afterstate tic-tac-toe
8. figure 8.8