Assignment1

[ratnasar]

import pip install gym
import numpy as np

Membuat lingkungan CartPole

env = gym.make('CartPole-v1')

Inisialisasi tabel Q dengan nilai-nilai acak

state_space_size = env.observation_space.shape[0]
action_space_size = env.action_space.n
q_table = np.random.rand(state_space_size, action_space_size)

Hyperparameters

learning_rate = 0.1
discount_factor = 0.99
exploration_prob = 1.0
exploration_decay = 0.995
max_episodes = 1000

for episode in range(max_episodes):
state = env.reset()
done = False
total_reward = 0

while not done:
    # Pilih tindakan dengan epsilon-greedy policy
    if np.random.rand() < exploration_prob:
        action = env.action_space.sample()  # Eksplorasi
    else:
        action = np.argmax(q_table[state, :])  # Eksploitasi
    
    next_state, reward, done, _ = env.step(action)
    
    # Update nilai Q dengan Q-learning
    q_table[state, action] = (1 - learning_rate) * q_table[state, action] + \
                             learning_rate * (reward + discount_factor * np.max(q_table[next_state, :]))
    
    state = next_state
    total_reward += reward

# Kurangi nilai epsilon (exploration)
exploration_prob *= exploration_decay

print(f"Episode {episode + 1}, Total Reward: {total_reward}")

Akhir dari pelatihan, Anda dapat menguji agen Anda di sini