Reinforcement-learning-agent

This repository contains a collection of different reinforcement learning projects, each implementing RL agent.

Each directory in this repository represents a distinct project, with its own agent and corresponding infrastructure. I have designed and implemented each of these projects.

Projects

• Actor-Critic: Implements an Actor-Critic reinforcement learning agent.
• DQN-experience-replay: Implements a Deep Q-Network (DQN) agent with experience replay.
• DynaQ-Qlearning: Implements a combination of Dyna-Q and Q-Learning.
• PPOKL: Implements a Proximal Policy Optimization (PPO) agent using Kullback-Leibler (KL) divergence.
• Rld: An older project implementing a unique reinforcement learning agent.
• TME1env: Contains an environment and corresponding agent for teaching purposes.
• ddpg: Implements a Deep Deterministic Policy Gradient (DDPG) agent.
• dqn-igs-her: Implements a DQN agent using Importance Sampling (IS) and Hindsight Experience Replay (HER).
• gail: Implements a Generative Adversarial Imitation Learning (GAIL) agent.
• gan: Implements a Generative Adversarial Network (GAN) agent.
• gridworld-random-agent: Implements a random agent in a Gridworld environment.
• multi-agent: Contains a multi-agent reinforcement learning project.
• variational-encoder: Implements a reinforcement learning agent with a Variational Autoencoder (VAE).

Feel free to explore each project's individual directory for more information about its structure and agent.

Note: .DS_Store is a system file used by the macOS operating system and does not contain any project-related content.