Algorithm 2 Pseudo code for learning algorithm for the selection of state action, policies, and rewards (Code Listing 2: Deep Q-learning algorithm)

1 Initialize P (state, action) with some random value

2 while (P! = terminal){

3     Initialize state

4 While (state! = terminal){

5     Choose an action from the state by policy inferred from P

6     Take action to action, observer, state’

7     P (state, action) ← P (state, action) + $\alpha$ [r + $\gamma$ max$\alpha$,

P (state’, action’) − P (state, action)]

8     state ← state’

9  }

10 }