Whats Worth Memorizing: Attribute-Based Planning for DEC-POMDPs

Abstract

Current algorithms for decentralized partially observable Markov decision processes (DEC-POMDPs) require a large amount of memory to produce high quality plans. To combat this, existing methods optimize a set of finite-state controllers with an arbitrary amount of fixed memory. While this works well for some problems, in general, scalability and solution quality remain limited. As an alternative, we propose remembering some attributes that summarize key aspects of an agents action and observation history. These attributes are often simple to determine, provide a well-motivated choice of controller size and focus the solution search on important components of agent histories. We show that for a range of DEC-POMDPs such attribute-based representation improves plan quality and scalability.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 18, 2008
Accession Number
AD1006340

Entities

People

  • Christopher Amato
  • Shlomo Zilberstein

Organizations

  • University of Massachusetts

Tags

Communities of Interest

  • Autonomy
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Algorithms
  • Artificial Intelligence
  • Autonomous Agents
  • Computer Programming
  • Computer Science
  • Dynamic Programming
  • Dynamics
  • Linear Programming
  • Machine Learning
  • Multiagent Systems
  • Nonlinear Programming
  • Observation
  • Probability
  • Probability Distributions
  • Scalability
  • Transitions
  • Uncertainty

Readers

  • Mathematical Modeling and Probability Theory.
  • Systems Analysis and Design
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.