Multiagent Planning Under Uncertainty With Communication
Speaker (s):
Matthijs Spaan
Assistant Professor of Computer Science
Delft University of Technology, the Netherlands
|
|
Date:
Time:
Venue:
|
|
June 9, 2016, Thursday
11:00am - 12:00pm
Meeting Room 4.4, Level 4
School of Information Systems
Singapore Management University
80 Stamford Road
Singapore 178902
We look forward to seeing you at this research seminar.
|
|
About the Talk
We consider decision-theoretic approaches to multiagent planning under
uncertainty, formalized as extensions of the Partially Observable
Markov Decision Process (POMDP) model. In cooperative settings,
communication between agents has the potential to significantly
improve team performance. For instance, a higher degree of
coordination can often be obtained by sharing local information. A
common objective, however, is to minimize the level of communication
required for satisfying performance. In this talk, we discuss
different models of communication that have been explored and pair
them with appropriate solution techniques. A particular focus is on
multi-robot planning problems, in which quality restrictions on the
communication network need to be taken into account.
About the Speaker
Matthijs Spaan is an assistant professor of Computer Science at Delft
University of Technology, the Netherlands. He holds a PhD degree in
Computer Science (2006) and an MSc degree in Artificial Intelligence
(2002), both from the University of Amsterdam. After graduating, he was
a research scientist at Instituto Superior Tecnico (Lisbon, Portugal)
and held a Marie Curie fellowship at Delft University of Technology.
His research focus has been on designing algorithms for planning under
uncertainty. In particular, he addresses the challenge of developing
intelligent agents using models like the Partially Observable Markov
Decision Process (POMDP) and its multiagent extensions. He has applied
his algorithms in several application domains such as robotics, smart
energy systems as well as transportation and traffic.
