TDDD10 |
AI Programming, 6 ECTS credits.
/AI-programmering/
For:
CS
D
DAV
IT
U
|
|
Prel. scheduled
hours: 64
Rec. self-study hours: 96
|
|
Area of Education: Technology
Main field of studies: Computer Science, Computer Engineering, Information Technology
|
|
Advancement level
(G1, G2, A): A
|
|
Aim:
The aim of the course is to give an introduction to agent-oriented AI
programming through the use of robotic soccer (RoboCup).
The course introduces fundamental problems and techniques related to
construction of agent-oriented AI-systems. The students should gain
practical experience in solving this kind of problems, as well as implementing and integrating these techniques.
The course has two phases: Introduction to AI programming using
RoboCup and RoboSoc, and in depth AI programming.
The aim of the first phase is to make the students proficient with
RoboCup and the RoboSoc framework which is used to implement the
agents. The goals of the first phase are that each student should:
- be able to use existing frameworks to implement simulated RoboCup agents, and
- be able to design and implement a multiagent system in the form of a RoboCup team.
The first aim of the second phase is to give an overview of problems
and solutions in the area of multiagent systems. The second aim is to
let each student focus on a specific problem or technique in order to
make an in depth study of it. The goals of the second phase are that
each student should:
- be able to list and explain important problems and techniques in the area of multiagent systems,
- be able to evaluate a technique or solutions to a problem in the area of multiagent systems, including summarizing and critizing existing work in order to make a judgement on the applicability or suitability of the chosen technique or solution,
- be able to design, implement, and test a multiagent technique in a simulated agent environment as part of a group, and
- be able to make written and oral presentations of their work.
|
|
Prerequisites: (valid for students admitted to programmes within which the course is offered)
Introductory course in artificial intelligence. Object-oriented programming, preferably in C++.
Note: Admission requirements for non-programme students usually also include admission requirements for the programme and threshhold requirements for progression within the programme, or corresponding.
|
|
Organisation:
The course is organised as a number of lectures and seminars combined
with practical work. The lectures discuss the practical and
theoretical aspects of agent-based AI-systems. The seminars discuss
specific solutions to problems found during laboratory assignments and
should make the connection between theory and practice stronger. The
seminars are held in cooperation with the students, who are expected
to present their problems and solutions. In the laboratory work the
students will implement and evaluate an AI technique and also develop
their own RoboCup team.
The course runs over the entire autumn semester.
|
|
Course contents:
Implementation and use of multi-agent systems, agent architectures,
machine learning, communication, teamwork and coordination among
agents.
|
|
Course literature:
Wooldridge, Michael (2002) An Introduction to Multiagent Systems. John Wiley & Sons. ISBN 047149691X.
|
|
Examination: |
|
Laboratory work |
6 ECTS
|
|
|
|
Course language is English.
Department offering the course: IDA.
Director of Studies: Peter Dalenius
Examiner: Cyrille Berger
Course Syllabus in Swedish
|