Postdoc at Southern California

Postdoctoral Research Associate Position
The Teamcore Research Group 
Computer Science Department 
University of Southern California 
July 2010

The Teamcore group (http://teamcore.usc.edu) is focused on research on multi-agent systems where multiple agents (including software agents, robots and people) may interact. We focus on fundamental research issues in Decision Theoretic (distributed MDPs), Game Theoretic, Belief-Desire-Intentions (BDI) and Distributed Constraint Reasoning (DCR) approaches to multiagent systems. In addition to fundamental research, our group is particularly interested in use-inspired research: applying research in multiagent systems to real-world problems and advancing fundamental research to address these needs. For example, our algorithms have been deployed for real-world use by several agencies including the LAX police, the Federal Air Marshals service and the Transportation security administration. We have a new opening for a post-doctoral research associate position starting in September 2010. Research will focus on the areas of fundamental research outlined above with emphasis on new algorithms and but also on their practical implementations / applications. Interested applicants should send their CV and have three letters of recommendation forwarded to Prof. Milind Tambe (tambe@usc.edu).

AAMAS Proceedings

The AAMAS 2010 proceedings are now available online.

Shohan and Leyton-Brown Textbook Free PDF

Somehow I had missed that you can download a free PDF of the Shoham and Leyton-Brown multiagent systems textbook. There you go, just in time for summer poolside reading season.

Multiagent Systems Textbook

I am teaching Multiagent Systems this semester and thus updating the textbook as my students find bugs in it. The latest "live" version is below, or you can just download the pdf. I also hope to update the netlogo models to Netlogo 4.1.

Market and Social Systems Engineering Program

The University of Pennsylvania will be starting a new program in market and social systems engineering. It seems to be geared towards a subset of multiagent researchers, those more interested in incentives and in building human-computer systems:

“Traditional programs don’t prepare students to design systems that take into account the goals and incentives of the people who use them,” said Michael Kearns, professor in the Department of Computer and Information Science in Penn’s School of Engineering and Applied Science and the program’s founding faculty director. “We haven’t asked engineering students to take a course in game theory to understand how incentives work or in sociology to understand human behavior. There is now enough science out there on the intersection of these topics to design undergraduate courses.”

Find out more by reading their press release and browsing the MKSE website.

Networks, Crowds, and Markets Book

The book titled Networks, Crowds, and Markets by David Easley and Jon Kleinberg is now available to download freely from the link above. The current version is a pre-publication draft.

As is evident from reading the outline, there is a lot of overlap between the research present in this book and multiagent research. They both deal with game theory, auctions, markets, bargaining and, most importantly, search in networks, albeit from different perspectives.

AAMAS facebook group

There is now a AAMAS facebook group. Go and try to post something funny about multiagent systems, I challenge you.

Algorithmic Game Theory Classes

Via the AGT blog, here is a list of classes about algorithmic game theory and multiagent systems.

Aroxo

Looks like Nick Jennings used his expertise in automated negotiation to help build Aroxo, a web application that lets agents negotiate a deal for you. More info at this BBC news article.

E-Commerce Conference Papers

The Games and Economic Behavior journal has a special issue on the 8th ACM Conference on Electronic Commerce, edited by David C. Parkes and Moshe Tennenholtz.

IJCAI Proceedings

The IJCAI 09 proceedings are online. A report from Ariel Proccacia highlights the 7 sessions they had on Algorithmic Game Theory.

I like Ariel's definition of AI as: "everything that might get published in AAAI/IJCAI", its a bit more specific that the old "once we understand it, it stops being AI" while still allowing the field to move around randomly.

AAMAS proceedings

AAMAS proceedings are freely available here.

Opera Unite

The new opera unite implements a web server into every web browsers, thus allowing real peer-to-peer communications, that is, we can build multiagent systems that run solely on the browser without the need for a server to route all messages. I hope the big three (IE, firefox, chrome) also implement this.

Algorithmic Game Theory Blog

Demonstrating how out of the loop I've been lately, I've just discovered a new blog by Noam Nissan on algorithmic game theory. Apparently, its been around for a couple of months. In any case, the book (Algorithmic Game Theory) is an excellent introduction to many of the problems that we face when building multiagent systems, from a formal perspective: meaning that the book focuses on theorems and algorithms. I'm hoping to soon been able to teach a class using both my textbook and AGT, to see how we can implement these algorithms.

Anyway, go check it out.

Fundamentals of Multiagent Systems with NetLogo Examples: March 2009

I have updated the textbook. You will find the new version below. The changes are all bugfixes. I am extremely thankful to the many people, from across the world, who have taken time to email me about the errors they have found it the book. I am also very happy that so many are reading it. Unfortunately, since I have not had the chance to teach our multiagent systems class lately, I have not had the time to add new material to the textbook. Hopefully, I will be teaching the class again in the next couple of years and then I will add new material. Still, I do fill confident that the material in the textbook is mostly timeless (for example, the Nash bargaining solution goes back to 1950) so it should hold up fine as a good introduction to the field.

I would be amiss if I didn't mention that Shohan and Leyton-Brown published Multiagent Systems: Algorithmic, Game-Theoretic, and Logical Foundations which covers much of the same material my book is trying to cover. Still, their coverage is not identical. Their book also focuses more on proofs. My goal, on the other hand, is to develop a book for practitioners: people who want to engineer multiagent systems. Thus, I try to focus on algorithms and provide the hands-on NetLogo examples. Anyway, you be the judge.

Fundamentals of Multiagent Systems

Publish at Scribd or explore others: Technology-Computer- multiagent