10-734 Foundations of Autonomous Decision Making under Uncertainty

12 units, Carnegie Mellon University

AI is increasingly being used not only for prediction but for decision making in the real-world. Algorithms for autonomous decision making need to rely not only the accuracy of predictions but also the uncertainty of predictions, to account for potentially good decisions that may not have been tried. This course will cover foundations of AI algorithms used for making decisions in the face of uncertainty starting from stochastic experimental design and Gaussian process optimization to advanced methods for sequential decisions such as bandits, online learning, active learning, and reinforcement learning. We will discuss these methods, analysis techniques, their sample and computational complexities, as well as open challenges related to these methods.

Prerequisites: (10-701 (min C) or 10-715 (min C)) and 10-716 (min C)

Offered: Fall 2025

Courseweave needs JavaScript to show sections, workloads and planning tools.