Reinforcement Learning

This repository is collection of reinforcement learning algorithms. At this moment, mostly classical techniques are included in this repository and hopefully will be updated with more recent state-of-the-art techniques.

Contents

RL Algorithms

1. Q-learning

2. SARSA

3. SARSA Lambda

4. R-learning

5. Policy Gradient - pending update

6. Deep Q-Network

7. Double DQN

8. Dueling DQN

9. Dueling Double DQN - pending update

10. Actor Critic - pending update

11. Deep Deterministic Policy Gradient - pending update

Applications & Experiments

1. Multi-Armed Bandit Problem

2. Stock Trading

3. OpenAI Gym Toy text environment Taxi-v3 | Agents | Overall Mean Scores | Time Taken(s) | | ------------- | ---------------------- | -------------- | |SARSA | -357.88 | 194.05 | |SARSA Lambda | -355.83 | 214.98 | |Q-Learning | -340.10 | 238.96 | |R-Learning | -339.89 | 293.08 |

4. Multi-Objective Problem in DST