Q-Learning Algorithm

Q-Learning Algorithm

1. For each $s,a$ set $\hat{Q}(s,a) = 0$.
2. Observe the current state $s$.
3. Select an action $a$ and execute it.
4. Receive reward $r$.
5. Update the table entry for $\hat{Q}(s,a)$ with \[ \hat{Q}(s,a) \leftarrow r + \gamma \max_{a'}\hat{Q}(s',a') \]
6. $s \leftarrow s'$ //Observe the new state.
7. Goto 3.