MIPC 2015 — Workshop at AAAI 2015

This workshop focuses on models and algorithms for multiagent interaction without prior coordination (MIPC). Interaction between agents is the defining attribute of multiagent systems, encompassing problems of planning in a decentralized setting, learning other agent models, composing teams with high task performance, and selected resource-bounded communication and coordination. There is significant variety in methodologies used to solve such problems, including symbolic reasoning about negotiation and argumentation, distributed optimization methods, machine learning methods such as multiagent reinforcement learning, etc. The majority of these well studied methods depends on some form of prior coordination. Often, the coordination is at the level of problem definition. For example, learning algorithms may assume that all agents share a common learning method or prior beliefs, distributed optimization methods may assume specific structural constraints regarding the partition of state space or cost/rewards, and symbolic methods often make strong assumptions regarding norms and protocols. In realistic problems, these assumptions are easily violated — calling for new models and algorithms that specifically address the case of ad hoc interactions. Similar issues are also becoming increasingly more pertinent in human-machine interactions, where there is a need for intelligent adaptive behaviour and assumptions regarding prior knowledge and communication are problematic.

Effective MIPC is most likely to be achieved as we bring together work from many different areas, including work on intelligent agents, machine learning, game theory, and operations research. For instance, game theorists have considered what happens to equilibria when common knowledge assumptions must be violated, agent designers are faced with mixed teams of humans and agents in open environments and developing variations on planning methods in response to this, etc. The goal of this workshop is to bring together these diverse viewpoints in an attempt to consolidate the common ground and identify new lines of attack.

This workshop is a successor of MIPC 2014, which was part of AAAI 2014.


The workshop will discuss research related to multiagent interaction without prior coordination, as outlined in the workshop description above. A non-exclusive list of relevant topics includes:

  • Learning and adaptation in multiagent systems without prior coordination
  • Agent coordination and cooperation without prior coordination
  • Team formation and information sharing in ad hoc settings
  • Teammate/opponent modelling and plan recognition
  • Human-machine interaction without prior coordination
  • Game theory/incomplete information applied to ad hoc agent coordination

Important Dates

  • Submission deadline: October 14, 2014
  • Extended submission deadline: November 9, 2014
  • Notification of acceptance: November 23, 2014
  • Camera-ready copies: December 1, 2015
  • Workshop: January 26, 2015

Submission Details

  • Papers can be submitted by November 9, 2014 to EasyChair:
    Note: Please do not submit your paper if you flagged it for this workshop in the AAAI submission. It will be passed on to us automatically and we will contact you on the notification date.
  • The workshop follows the formatting guidelines for standard paper submissions to the AAAI-15 main track. See here for details:
  • Papers will be selected based on a peer review process. Note that submissions are not anonymous anymore, please include your names and affiliations.

Talk-Only Option

This year, we offer a talk-only option for authors of relevant papers that have been published in journals or conference proceedings. Interested authors are encouraged to send their paper (in PDF or PS format) and publication details via e-mail to mipc2015 AT If the paper is deemed relevant for the workshop, we will allocate a presentation slot for the authors in the workshop program.

Accepted Papers


Monday 26 January — Room 'Hill Country C' — Hyatt Regency, Austin

09:00 - 10:30Session 1 (Chair: Stefano Albrecht)
09:00 - 09:15Opening Remarks
09:15 - 09:40William Curran, Adrian Agogino, Kagan Tumer:
Agent Partitioning with Reward/Utility-Based Impact
09:40 - 10:05Katie Genter, Shun Zhang, Peter Stone:
Placing Influencing Agents in a Flock
10:05 - 10:30Alessandro Panella, Piotr Gmytrasiewicz:
Nonparametric Bayesian Learning of Other Agents' Policies in Interactive POMDPs
10:30 - 11:00Coffee Break
11:00 - 12:30Session 2 (Chair: Jacob Crandall)
11:00 - 11:40Invited Talk — Pascal Poupart:
Leveraging Expert Feedback in Recommender Systems
11:40 - 12:05Stefano Albrecht, Jacob Crandall, Subramanian Ramamoorthy:
E-HBA: Using Action Policies for Expert Advice and Agent Typification
12:05 - 12:30Trevor Sarratt, Arnav Jhala:
RAPID: A Belief Convergence Strategy for Collaborating with Inconsistent Agents
12:30 - 14:00Lunch
14:00 - 15:30Session 3 (Chair: Stefano Albrecht)
14:00 - 14:40Invited Talk — Sarit Kraus:
Agent-Human Interaction without Prior Communication
14:40 - 15:05Sayan Sen, Julie Adams:
Real-time Optimal Selection of Multirobot Coalition Formation Algorithms using Conceptual Clustering
15:05 - 15:30Katie Genter, Tim Laue, Peter Stone:
The RoboCup 2014 SPL Drop-in Player Competition: Experiments in Teamwork without Pre-coordination
15:30 - 16:00Coffee Break
16:00 - 17:00Session 4 (Chair: Jacob Crandall)
16:00 - 16:40Invited Talk — Piotr Gmytrasiewicz:
Interactive POMDPs
16:40 - 17:00Closing Remarks

Invited Talks

Pascal Poupart
David R. Cheriton School of Computer Science
University of Waterloo

Leveraging Expert Feedback in Recommender Systems

Machine learning offers numerous methods to learn the parameters of recommender systems based on past data. However, in some applications such as fault diagnostics, data correspond to rare events (since faults are not supposed to be frequent) and therefore there may not be enough data to learn the parameters. In this talk, I will explain how to leverage expert feedback to iteratively refine recommender systems modeled as Bayesian networks and Markov decision processes. This will be done without explicit coordination with the expert by taking advantage of implicit constraints that arise as the expert interacts with the recommender system. The approach will be demonstrated with diagnostic sessions from a manufacturing scenario.

Sarit Kraus
Department of Computer Science
Bar-Ilan University

Agent-Human Interaction without Prior Communication

How should an agent make strategic decisions with people when they have no prior history of interaction? Examples of such situations can include first-time (or anonymous) users in e-commerce systems and navigation systems in rental cars. In these situations it is crucial for an automated agent to be able to adapt such that people are satisfied with the interaction and where the agent makes minimal compromises. A natural approach is to use Machine Learning to model a general population and then use the model to adapt to the individual person. I will identify several challenges to this approach, including the problem of sparse data, personality, complex environments and culture, and show ways to address them.

Piotr Gmytrasiewicz
Department of Computer Science
University of Illinois at Chicago

Interactive POMDPs

We will introduce a theoretical framework of Interactive Partially Observable Markov Decision Processes (IPOMDPs), which generalize classical POMDPs for multi-agent environments. We will motivate the framework and illustrate its basic assumptions, as well as it limitations. We will give examples and contrast results with other frameworks. We will conclude with challenges of future research directions.

Program Committee

  • Noa Agmon (Bar-Ilan University)
  • Bo An (Nanyang Technological University)
  • Brian Coltin (NASA Ames Intelligent Robotics Group)
  • Sam Devlin (University of York)
  • Christopher Geib (Drexel University)
  • Shivaram Kalyanakrishnan (Indian Institute of Science)
  • Gal Kaminka (Bar-Ilan University)
  • Balajee Kannan (General Electric)
  • Marc Lanctot (Google DeepMind)
  • Robert Loftin (North Carolina State University)
  • Bryan Low (National University of Singapore)
  • James MacGlashan (Brown University)
  • Stefanos Nikolaidis (Massachusetts Institute of Technology)
  • Benjamin Rosman (CSIR South Africa)
  • Michael Rovatsos (University of Edinburgh)
  • Abdallah Saffidine (University of New South Wales)
  • Aris Valtazanos (University of Edinburgh)
  • Logan Yliniemi (Oregon State University)
  • Dongmo Zhang (University of Western Sydney)


Program chairs: Advisory committee:


If you have any questions about the MIPC workshop, please contact the organizers at:
mipc2015 AT