Problem Solving Techniques in Artificial Intelligence

Course Syllabus

Course Number: 181.147 - Lecture with Exercises (SS 2006)

Course Title: AK der AI 2: Problem Solving Techniques in Artificial Intelligence

Credits: 2,0 h (3 ECTS)

Schedule: Wednesday 16:00 – 18:00 (first class: 27.04.2005, 17:00-19:00) in Seminarraum 184/2

Lecturer: Nysret Musliu

Email: musliu@dbai.tuwien.ac.at

If you want to visit this course, please register it in TUWIS++

Course Description:

This course covers most important artificial intelligence search techniques used for problem solving in past years. The emphasis will be in heuristic techniques for constraint satisfaction problems and advanced heuristic search techniques, like tabu search and genetic algorithms. The topics covered by this course will be:

· Basic concepts of problem solving

· Traditional methods

· Constraint satisfaction problems

· Heuristics for constraint satisfaction problems

· Local search

· Stochastic hill climbing

· Simulated annealing

· Tabu search

· Genetic algorithms

The use of heuristic methods will be illustrated by examples, which show the application of these methods for solving of different hard problems with high practical relevance. Moreover, successful use of heuristic methods in industrial project developed by our Institute will be presented.

Course Structure:

This course is combination of lectures and exercises. Exercises part consists on small project, which will be an implementation of one of heuristic methods for particular problem. The problem for which the heuristic method will be implemented can be selected independently from student. Additionally, during the course discussion for solving of different logical problems and puzzles will be made. Proposals of students for discussions about problems for which they are interested are encouraged.

Literature:

Z. Michalewicz and D. B. Fogel. How to Solve It: Modern Heuristics. Springer-Verlag. 2000.

Stuart Russell and Peter Norvig . Artificial Intelligence: A Modern Approach (Second Edition), 2003

Additionally the key papers for heuristics from the literature will be used.

…

Schedule of Classes

27.04.2006 (17:00): Problem Solving: Basic Concepts

27.04.2006 (17:00): Traditional methods

03.05.2006 (16:00): Constraint Satisfaction (Constraint Satisfaction Problems, Constraint propagation)

10.05.2006 (16:00): Constraint Satisfaction (Hypertree Decompositions, Heuristic for Constrain Satisfaction Problems)

17.05.2006 (16:00): Local Search: Hill Climbing and Simulated Annealing

17.05.2006 (16:00): Local Search: Tabu Search

24.05.2006 (16:00): Local Search Techniques Applications

31.05.2006 (16:00): Genetic Algorithms

07.06.2006 (16:00): Genetic Algorithms

15.06.2006 (16:00): Constraint-Handling Techniques

21.06.2006 (16:00): Project Presentations & Final Exam

mailto: musliu@dbai.tuwien.ac.at