EE550 Applied Deep Reinforcement Learning
EE550 Applied Deep Reinforcement Learning
Embedded Files
This is a public information for the course EE550 Applied Deep Reinforcement Learning started from Winter 2025 at the Department of Electrical and Computer Engineering at the Royal Military College of Canada. Enrolled students should consult the course website hosted on Moodle.
Course Description
This course provides a comprehensive exploration of deep reinforcement learning (DRL), an exciting field at the intersection of artificial intelligence, machine learning, and control systems. The materials of this course combine both theory and practice while covering the major methods of DRL.
The studied materials include the concept of reinforcement learning, Markov decision processes, elements of a reinforcement learning environment, Q-functions, value-based RL methods, policy-based RL methods, action space exploration and exploitation, and Inference. The course also sheds the light on advanced topics such as inverse reinforcement learning, meta-learning, and transfer learning.
The course will provide students with a deep understanding of foundational principles, algorithms, and practical applications of reinforcement learning. It will also expand their skills in using deep reinforcement learning to solve real-world decision-making problems. This part will be achieved through a project.
Main Topics
Reinforcement Learning Intro
Markov Decision Process (MDP)
OpenAI Gymnasium with Python
Deep Learning with PyTorch
Value-based methods and Deep Q-networks (DQNs)
Experience Replay
Policy-based methods and Actor-Critic algorithms
Optimal Control and Planning
Advanced Topics: Inverse Reinforcement Learning
Advanced Topics: Transfer learning, and Meta-learning
Project
The aim of the project is to explore meaningful ideas in deep reinforcement learning for creating useful tasks and problem-solving approaches. The students can work on a topic related to their research area. They are also welcome to use the robotic arm shown in the next figure to demonstrate some functionality.
The project will consist of three stages:
Abstract and Literature Review
Progress Report
Final Report
The project will be evaluated based on the following:
Quality and novelty of idea
Technical soundness
Logical progression
Report and presentation quality
Skills demonstration
Results, analysis and meaningful exploration
Yahboom JetCobot 7-axis visual collaborative robotic arm-Jetson NANO 4GB which is equipped by a camera.
Textbook
We will use the following textbooks for the course:
Main textbook:
Maxim Lapan, "Deep Reinforcement Learning Hands-On: A practical and easy-to-follow guide to RL from Q-learning and DQNs to PPO and RLHF". Third Edition, Packt Publishing Ltd, UK, October, 2024. Link
Zhao, Shiyu. Mathematical Foundations of Reinforcement Learning. Springer Verlag, Singapore, 2025. Link
Supplementary textbook: Richard S. Sutton and Andrew G. Bartow, "Reinforcement Learning: An Introduction", Second Edition, MIT press, 2020. http://incompleteideas.net/book/RLbook2020.pdf
Page updated
Google Sites
Report abuse