Python: Advanced Guide to Artificial Intelligence

Q-Learning with Q-Table

You may follow the code for this section in ch-13b.ipynb. Since our discretised space is of the dimensions [10,10,10,10], our Q-Table is of [10,10,10,10,2] dimensions:

# create a Q-Table of shape (10,10,10,10, 2) representing S X A -> Rq_table = np.zeros(shape = np.append(n_s,n_a)) 

We define a Q-Table policy that exploits or explores based on the exploration_rate:

def policy_q_table(state, env):    # Exploration strategy - Select a random action    if np.random.random() < explore_rate:        action = env.action_space.sample()    # Exploitation strategy - Select the action with the highest q    else:        action = np.argmax(q_table[tuple(state)])    return action

Define the episode() function that runs a single episode as follows:

1. Start with ...