resulemreaygan / reinforcement-learning-project

Deep Learning Türkiye - Reinforcement Learning Project

This repository consists projects from Deep Learning Türkiye - Reinforcement Learning Group. Enter folders to see each project's details.

1. Introduction To RL

Simple tic tac toe example. Learns via Value Function at the moment. Policy Search TODO. Benefited from tansey.

2. Multi-Armed Bandits

Provides the underlying testbed for bandit problem.

3. Finite Markov Decision Processes

Uses the OpenAI Gym. Learns via Q-Learning.

4. Temporal Difference

Multiple approaches to CartPole problem. Benefited from dennybritz.

Library usage

You can find example usage below.

import gym
from lib import q_learning_agent, double_q_learning_agent, sarsa_learning_agent

env = gym.make("FrozenLake-v0")
env.reset()

def train(agent):
    for i_episode in range(1000):
        state = env.reset()
        while True:
            action = agent.select_action(state)
            next_state, reward, done, _ = env.step(action)
            agent.learn(action, reward, state, next_state)
            if done:
                break
            state = next_state

qla = q_learning_agent(epsilon=0.3, discount_factor=0.9, alpha=0.5, action_space=env.action_space.n)
sla = sarsa_learning_agent(epsilon=0.3, discount_factor=0.9, alpha=0.5, action_space=env.action_space.n)
dqla = double_q_learning_agent(epsilon=0.3, discount_factor=0.9, alpha=0.5, action_space=env.action_space.n)

train(qla)
train(sla)
train(dqla)