Teaching Summer 2025

• Models and Solvers in AI
• State models and heuristic search
• Logic, SAT solving, and ASP solving
• Classical planning: language, model, basic algorithms (heuristic search and SAT)
• Markov Decision Processes (MDPs): basic models and algorithms (VI, PI, RTDP, MCTS)
• Deep learning (DL) as model and solver; supervised learning, Approx VI, PI, MPI with DL
• Reinforcement Learning (RL): Value-based and policy gradient methods
• Current research in planning and RL

This course explores the core challenges of mobile robotics, focusing on navigating uncertainty. We begin with probability theory fundamentals, essential for state estimation and mapping. Students will explore localization techniques, including Markov and Monte Carlo methods, and tackle Simultaneous Localization and Mapping (SLAM). The course then transitions to decision-making under uncertainty, covering Markov Decision Processes (MDPs) and Partially Observable Markov Decision Processes (POMDPs), and studies reasoning about actions in unpredictable environments.