About me

I am a Postdoctoral Associate in the learning and intelligent systems group (LIS) at MIT. I work together with Leslie Kaelbling on a project on multiagent reinforcement learning. I obtained my PhD at the university of Amsterdam under the guidance of Frans Groen and Nikos Vlassis.

As a researcher in computer science (CS) and artificial intelligence (AI) I try to generate fundamental knowledge about algorithms and models for complex tasks. At the same time I try to think about applications for such techniques. That way, I hope to be able to contribute to technological innovations and focus my research on areas with potential impact for the real world.

My main interests lie in the field of sequential decision making under uncertainty. This field studies methods that allow an intelligent system, an agent, to make decisions over time. I am particularly interested in multiagent systems: settings with multiple agents. Sequential decision making is studied in, and has close ties to many research fields within AI, CS and economics such as: planning, reinforcement learning, optimization, machine learning, regression, classification, graphical models, probabilistic inference and game theory.

My PhD research focused on multiagent decision making under uncertainty In particular, I investigated settings in which a team of agents has to perform a task without communication and in which the agents cannot observe the true state of their environment (i.e., the environment is partially observable). Such situations can be modeled by decentralized POMDPs (Dec-POMDPs) and partially observable stochastic games (POSGs).

News

• Two of my papers will appear at AAMAS: a full paper about heuristic search of multiagent influence and an extended abstract about effective method for performing the vector-based backup in settings with delayed communication.
• Our workshop proposal for next year's MSDM has been accepted, so make sure to start preparing your papers!
• The MIT news office wrote about my work! Read about it here.
• I finished my book chapter on decentralized POMDPs. It can be used as a introduction to Dec-POMDPs and also provides insight on how the current state-of-the-art algorithms (forward heuristic search and backward dynamic programming) for finite-horizon Dec-POMDPs relate to each other.
• The 6th MSDM workshop was a great succes! We had 12 strong accepted papers. Check the proceedings at the MSDM website.
• Our paper Scaling Up Optimal Heuristic Search in Dec-POMDPs via Incremental Expansion got accepted at IJCAI'11. It presents a significant improvement over the state of the art in optimal solutions for Dec-POMDPs. Also check the MSDM version for more details.
• I am co-organizing MSDM 2011, the workshop on Multiagent Sequential Decision Making in Uncertain Domains. Deadline for submissions is Januari 31 2011. For more info, see the MSDM website.
• I gave a lecture on decision making for cooperative agents in Leslie's class on planning and decision making. Slides can be found under teaching.
• My new website is online - you're looking at it! For many pages, I'm still linking back to my old website at the University of Amsterdam.
• I've started at MIT on July 15th!
• Look my thesis is on google books!
• On February 12th 2010 I succesfully defended my thesis, titled "Value-Based Planning for Teams of Agents in Stochastic Partially Observable Environments".
• Prashant Doshi and his students Christopher Jackson and Kennth Bogert used the multiagent decision project (MADP) toolbox to compute policies in the game of StarCraft. Watch the (pretty cool) video.