Other years: [Spring 2003] [Spring 2001]

Topics of this course include theory, implementation technology and applications of agent-based systems.

• Registrations using TOPI or by attending the first two lectures.
• Lectures by D.Sc.(Tech.) Tomi Janhunen: Tuesdays 11-14, room TB353
• Tutorials and project assignments by M.Sc.(Tech.) Tommi Syrjänen: Thursdays 15-16, room TB353
• Course material:
  
  • Course book: S. Russel and P. Norvig: Artificial Intelligence: A Modern Approach, Prentice Hall, 1995.
  • Seminar: Selected articles (to be announced) from
    • Proceedings of the 17th Joint International Conference on Artificial Intelligence (IJCAI'01)
    • M. Huhns and M. Singh (Eds.): Readings in Agents, Morgan Kaufmann, 1997
• Project assignments will be related to the simulation league of RoboCup.
• In order to pass the course one has to
  • give a seminar talk
  • complete a project assignment and
  • pass an exam.
• A more detailed brochure in Finnish.

• May 7, 2002, final results (published on May 28, 2002)
• September 3, 2002, final results (published on Nov 26, 2002)
• January 8, 2003, final results (published on Apr 28, 2003)
• Please register for the exams using TOPI.

January 15: Introduction

Practical arrangements, a brief introduction (slides in ps and pdf)
Tutorial (January 17): RoboCup demo

January 22: Lecture 1

Intelligent agents [R&N: Chapter 2] (slides in ps and pdf)
Tutorial (January 24): project proposals

January 29: Lecture 2

Uncertainty [R&N: Chapter 14] (slides in ps and pdf)
Tutorial 1 (January 31), solutions

February 2: Lecture 3

Belief networks [R&N: Chapter 15] (slides in ps and pdf)
Tutorial 2 (February 7), solutions

February 12: Lecture 4

Making simple decisions [R&N: Chapter 16] (slides in ps and pdf)
Tutorial 3 (February 14), solutions

February 19: Lecture 5

Making complex decisions [R&N: Chapter 17] (slides in ps and pdf)
Tutorial 4 (February 21), solutions

February 26: Lecture 6

Learning [R&N: Chapters 18.1-18.5, 18.7, and 19.6] (slides in ps and pdf)
Tutorial 5 (February 28), solutions

• Are based on selected articles.
• You should make your choice (article & presentation time) by March 4, 2002, by sending an email to the lecturer.
• Requests are processed in the order they are received.
• Presentations should last 40 minutes, 5 minutes is reserved for questions and discussion.
• For results, please see the list of accepted presentations.

March 19: Session 1

11:15-12:00 Mikko Särelä: A Layered Brain Architecture for Synthetic Creatures by Damian Isla, Robert Burke, Marc Downie, and Bruce Blumberg

12:15-13:00 Antti Salovaara: Modeling Rational Agents within a BDI-Architecture by Anand Rao and Michael Georgeff

13:15-14:00 Heikki Rantanen: COLLAGEN: When Agents Collaborate with People by Charles Rich and Candace L. Sidner

March 26: Session 2

11:15-12:00 Riku Saikkonen: Agent-Oriented Programming by Yoav Shoham

12:15-13:00 Timo Lilja: Semantics and Conversations for an Agent Communication Language by Yannis Labrou and Tim Finin

13:15-14:00

April 2: No session (Easter vacation)

April 9: Session 3

12:15-13:00Pierre Moermans: A Multiagent System for Helping Urban Traffic Management by L.A. Garcia and F. Toledo

April 16: Session 4

11:15-12:00 Matti Järvisalo: Coordination without Communication: Experimental Validation of Focal Point Techniques by Maier Fenster, Sarit Kraus, and Jeffrey S. Rosenschein

12:15-13:00 Oleg Myrk: Global Information Management via Local Autonomous Agents by Michael N. Huhns, Munindar P. Singh, and Tomasz Ksiezyk

April 23: Session 5

11:15-12:00Jussi Rautio: Rational and Convergent Learning in Stochastic Games by Michael Bowling and Manuela Veloso

12:15-13:00 Maurizio Caltabiano: Multi-Agent Influence Diagrams for Representing and Solving Games by Daphne Koller and Brian Milch

• You should make a short project proposal (at most one page) by February 18.
• Proposals can be sent directly to Tommi Syrjänen for his approval.
• List of passed project assignments.

• The project team (from 1 to 3 persons).
• What is to be done:
  • Basically you should implement a soccer playing agent that is able to make a goal and to act as a goalie.
  • It is also possible to make enhancements to an existing implementation.
• What tools, programming language(s) and libraries are to be used.

Back to menu.