IRL & Online Program
Applied Reinforcement Learning
Train autonomous agents for gaming, robotics, and complex decision-making systems using modern RL algorithms.
Hands-on Highlights
- Implement Q-Learning from scratch
- Train a Deep Q-Network (DQN) to play Atari games
- Apply PPO (Proximal Policy Optimization) to robotics simulations
- Build a multi-agent trading environment
Detailed Syllabus
Week 1-2
RL Fundamentals
- Markov Decision Processes (MDP)
- Value Iteration and Policy Iteration
- Exploration vs Exploitation (Epsilon-Greedy)
- Implementing Q-Learning from scratch
Week 3-5
Deep Q-Networks (DQN)
- Combining deep learning with Q-learning
- Experience Replay and Target Networks
- Solving OpenAI Gym environments
- Training an agent to play Atari games
Week 6-8
Policy Gradients & PPO
- The REINFORCE algorithm
- Actor-Critic architectures
- Proximal Policy Optimization (PPO) theory
- Applying PPO to continuous control (Mujoco)
Week 9-10
Multi-Agent RL & Advanced Topics
- Multi-Agent environments (PettingZoo)
- Reward shaping strategies
- Offline reinforcement learning
- Capstone project: Custom RL agent
Target Roles & Career Paths
Reinforcement Learning Researcher
Robotics AI Engineer
Game AI Developer
Autonomous Systems Engineer
These are the primary roles you will be equipped to apply for upon successful completion of the course and portfolio projects.