Learning Unknown Event Models

Abstract

Agents with incomplete environment models are likely to be surprised, and this represents an opportunity to learn. We investigate approaches for situated agents to detect surprises, discriminate among different forms of surprise, and hypothesize new models for the unknown events that surprised them. We instantiate these approaches in a new goal reasoning agent (named FOOLMETWICE), investigate its performance in simulation studies, and report that it produces plans with significantly reduced execution cost in comparison to not learning models for surprising events.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 01, 2014
Accession Number
ADA610456

Entities

People

  • David W. Aha
  • Matthew Molineaux

Organizations

  • Knexus Research (United States)

Tags

Communities of Interest

  • Autonomy
  • Space

DTIC Thesaurus Topics

  • Artificial Intelligence
  • Artificial Satellites
  • Autonomous Agents
  • Autonomous Underwater Vehicles
  • Autonomy
  • Cognitive Science
  • Differential Equations
  • Engineering
  • Environment
  • Learning
  • Machine Learning
  • Multiagent Systems
  • Reasoning
  • Reinforcement Learning
  • Simulations
  • Software Development
  • Training

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Computational Modeling and Simulation
  • Military History of the United States in the 20th Century.