G52PAS Planning and Search 2012-2013

Lecture slides

Lecture 1 Introduction, revision of basic search, definition of planning, plan of the module.
Lecture 2 revision of search continued.
Lecture 3 local search.
Lecture 4 genetic algorithms.
Lecture 5 search and SAT. Exercise . Solution to the exercise .
Lecture 6 Search with non-determinism and partial observability
Exercise on and-or search and answer.
30 October: Lecture 6 continued (on the 25th I only covered non-determinism, on the 30th I will do partial observability)
Lecture 7a Logical agents.
Lecture 7b First-order logic. Exercise on first-order logic . Answer to the exercise
Lecture 8 Situation calculus. Exercise on situation calculus . Answer to the exercise
Lecture 9 Classical planning: progression and regression planning.
Lecture 10 Classical planning: goal stack planning.
Exercise: finish the trace of the goal stack algorithm on the example from the lecture.
Two more exercises: trace the same example (Sussman anomaly) for progression planning using breadth first search and regression planning using backward BFS as in Lecture 9.
Lecture 11 Classical planning: partial order planning.
Exercise: see the lecture slides (trace a blocks world example with partial order planning).
Lecture 12 Classical planning: GraphPlan.
Lecture 13 Classical planning: SatPlan. Complexity of classical planning, comparison of classical planning systems performance. Exercise: please do question 3 from 2010 exam paper
Lecture 14 planning with time and resources
Lecture 15 HTN planning (planning with abstraction)
Lecture 16 planning and acting in non-deterministic domains.
Lecture 17 revision.

This file is maintained by Natasha Alechina
Last updated 19 February 2013.