Automated Planning (TDDD48, Master)

This course covers automated planning, a central topic in AI, with hands-on experience in creating planning models to solve sequential decision making problems, exploring its numerous applications from logistics to space exploration.

The course and the labs will be completely redesigned for 2024 to cover the latest research on AI planning. Here is the new list of planned topics:

• What is Planning?
  • Transition Systems and Propositional Logic
  • State Variables, Operators and Planning Tasks
  • Computational Complexity of Planning
• Overview of Classical Planning Algorithms
  • Progression and Regression Search
  • SAT Planning
  • Symbolic Search
• Heuristics
  • Delete Relaxation
  • Abstractions
  • Critical Path Heuristics
  • Landmarks
• Linear Programming Heuristics
  • Cost Partitioning
  • Post-hoc Optimization
  • Network Flow Heuristics
  • Operator Counting
  • Potential Heuristics
• Probabilistic Planning
  • Stochastic Shortest Path Problems
  • Markov Decision Processes
• Reinforcement Learning

Course Information

• Examiner: Jendrik Seipp
• Assistant: Mika Skjelnes
• Syllabus: TDDD48