AI-Enabled Sequential Decision Making

1 Unit · Graduate

Formulation and solution of sequential decision-making problems under uncertainty. Markov decision processes and exact solution methods, including value iteration, policy iteration, and linear programming. Model-based methods, including model predictive control and Monte Carlo tree search. Model-free reinforcement learning methods, including deep Q networks, policy gradient, and actorcritic algorithms. Students implement these methods in Python and apply them through case studies that highlight the capabilities and limitations of each approach.