Reinforcement Learning
Are you ready to explore the world of reinforcement learning (RL) and discover how intelligent systems make decisions?
This course will guide you from the fundamentals of RL—where agents learn and improve their behavior—right through to advanced, powerful techniques that drive real-world AI. You won’t just learn the “how” but also the “why” behind methods like Deep Q-Learning (DQN), Proximal Policy Optimization (PPO), and Actor-Critic models. These skills will let you create agents that can master games, navigate environments, and respond to changing surroundings.
With hands-on coding in PyTorch, you’ll apply these skills directly—building agents to balance poles, land lunar modules, and tackle real-time interactive challenges. By the end, you’ll have the confidence and knowledge to design, improve, and deploy your own intelligent agents.
Start this journey and unlock the potential to create smart, adaptable agents through the power of reinforcement learning!
Course Content
Introduction
Gynasium Environments for Reinforcement Lerning
Deep Reinforcement Learning
On Policy Methods
- Introduction to On-Policy Methods
- Policy Optimization
- Likelihood Ratio Policy Gradient
- Policy Gradient Estimation with Likelihood Ratios
- Understanding Baseline Subtraction in Policy Gradient Methods
- Baseline Choices in Policy Gradient Methods
- Value Function Estimation
- Vanilla Policy Gradient Algorithm
- Enhancing Policy Gradients with A3C and GAE
- Structured On-Policy Training