Code:

BERMASC013

Study year:

2007-2008

Long name:

Multi Agent Systems Research

ECTS:

5

Language:

English

Lecturer(s):

Dr. W. Ketter, prof. dr. ir. H.W.G.M. van Heck

Contact person:

Dr. W. Ketter

Coordinator:

Dr. W. Ketter

Faculty:

RSM Erasmus University

Number of lectures:

Workshop style – 3 full days at the beginning and 2 full days at the end

Hours per lecture:

5

Goal:

Expose students to the state of the art in research on multi-agent systems to gather information and facilitate decision making in business and economic environments. The course is designed to help students to develop and deepen their own research ideas and proposals.

Course contents:

This course is designed to familiarize Research Master’s and PhD students with a wide variety of issues in the domain of multi-agents systems. The study of agents presents a unique opportunity to integrate results from many diverse areas of research, such as artificial intelligence, behavioral science, computer science, economics, information systems, operations research, and software engineering. Thus the aim of this course is to expose students to the state of the art in research on multi-agent systems to gather information and to facilitate decision making in business and economic environments. In addition to providing students with knowledge in the area of multi-agents systems, students will get familiarized with the methods and paradigms used in the area. The course is designed to help students to develop and deepen their own research ideas.

Modern business networks and markets are highly dynamic and exhibit a high degree of uncertainty. Under these conditions business managers are routinely faced with complex strategic, tactical, and operational decisions; decisions ranging from the macroscopic (i.e. which markets should we enter and when?) to the microscopic (i.e. which products should be packed on which pallet?). Also customers are faced with multi attribute decisions, such as from whom can I book a travel under certain constraints (money, time, quality, etc)? Within this workshop we investigate how learning agents may be designed to support humans in these decision making processes. We define learning agents as software entities that carry out some set of operations on behalf of a user or another program with some degree of independence or autonomy, improve their performance from experience and in so doing employ some knowledge or representation of the user’s goals or needs.

This workshop provides a broad introduction to autonomous agents with an emphasis on multiagent systems. Topics include:

Agent architectures and modeling

Inter-agent communication and trust

Collective intelligence and cognitive collaboration

Teamwork and distributed rational decision making

Mechanism design and auctions

Multiagent learning

The second main emphasis of the workshop lies on applications of multi-agent system to gathering information and facilitate decision making. Topics may include:

· Supply-chain management

o Procurement, Production, Sales agents

· Electronic markets

o Auction markets (Flower auctions, eBay, …)

o Recommender agents (Shop bots, Dell, ...)

· Transportation

o Dynamic fleet management

· Mobile telecommunication

o Call routing and billing

Examination:

Contribute to the discussion of the assigned articles

Present papers to the rest of the class

Write a original research proposal due at the end of the course

Literature:

· Academic articles

· Recommended reading: "Multiagent systems. A modern approach to Distributed Artificial Intelligence." edited by Gerhard Weiss. The MIT Press, 1999. ISBN: 0262731312