MSc in Computing Science - Reinforcement Learning (Thesis-based) in Edmonton Canada | University of Alberta

Reinforcement learning encompasses theoretical frameworks and computational techniques for optimal decision-making, developed over the past quarter-century within machine learning and operations research, while also gaining significance in psychology and neuroscience. These approaches provide effective approximations for complex optimal-control challenges that prove too extensive or ambiguous for traditional methods like dynamic programming. Notable applications include autonomous helicopter flight, elevator control systems, backgammon gameplay, and scheduling with limited resources, where reinforcement learning has achieved state-of-the-art results.

At the University of Alberta, researchers focus on advancing reinforcement learning by overcoming current limitations to broaden its practical use, while also exploring its potential as an intelligence model capable of human-like performance. This work involves mathematical innovation, computational testing, robotic implementations, game strategy development, and creating computational simulations of natural learning mechanisms.