Reinforcement Learning Specialization (University of Alberta)

The Reinforcement Learning Specialization from the University of Alberta is a four-course online series that teaches the foundations and practical implementation of reinforcement learning (RL). The Specialization covers core RL topics including Temporal-Difference learning, Monte Carlo methods, Q-learning, policy gradient methods, function approximation, and culminates in a capstone project where learners implement a complete RL solution.

The series focuses on small-scale problems to build a solid theoretical and practical understanding applicable to game AI, IoT, clinical decision-making, industrial control, finance and more.

This is a self-paced online program delivered via Coursera. The Specialization comprises 17 modules across four courses and includes programming assignments and quizzes; a certificate is awarded on completion.

Recommended prerequisites include comfort programming in Python (able to convert pseudocode to Python), basic statistics, linear algebra and calculus. The program is aimed at intermediate learners (recommended at least one year of undergraduate CS or 2-3 years professional software development experience).