Peter Stone's Selected Publications

Classified by TopicClassified by Publication TypeSorted by DateSorted by First Author Last NameClassified by Funding Source

Design Principles for Creating Human-Shapable Agents

W. Bradley Knox, Ian Fasel, and Peter Stone. Design Principles for Creating Human-Shapable Agents. In AAAI Spring 2009 Symposium on Agents that Learn from Human Teachers, March 2009.
AAAI Spring 2009 Symposium: Agents that Learn from Human Teachers

Abstract

In order for learning agents to be useful to non-technical users, it is important to be able to teach agents how to perform new tasks using simple communication methods. We begin this paper by describing a framework we recently developed called Training an Agent Manually via Evaluative Reinforcement (TAMER), which allows a human to train a learning agent by giving simple scalar reinforcement\footnoteIn this paper, we distinguish between human reinforcement and environmental reward within an MDP. To avoid confusion, human feedback is always called reinforcement''. signals while observing the agent perform the task. We then discuss how this work fits into a general taxonomy of methods for human-teachable (HT) agents and argue that the entire field of HT agents could benefit from an increased focus on the \em human side of teaching interactions. We then propose a set of conjectures about aspects of human teaching behavior that we believe could be incorporated into future work on HT agents.

BibTeX Entry

@InProceedings{AAAIsymp09-knox,
author="W.\ Bradley Knox and Ian Fasel and Peter Stone",
title="Design Principles for Creating Human-Shapable Agents",
booktitle="AAAI Spring 2009 Symposium on Agents that Learn from Human Teachers",
month="March",
year="2009",
abstract={In order for learning agents to be useful to non-technical users, it
is important to be able to teach agents how to perform new tasks using
simple communication methods. We begin this paper by describing a
framework we recently developed called Training an Agent Manually via
Evaluative Reinforcement (TAMER), which allows a human to train a
learning agent by giving simple scalar reinforcement\footnote{In this
paper, we distinguish between human reinforcement and environmental
reward within an MDP. To avoid confusion, human feedback is always
called reinforcement''.} signals while observing the agent perform
the task. We then discuss how this work fits into a general taxonomy
of methods for human-teachable (HT) agents and argue that the entire
field of HT agents could benefit from an increased focus on the {\em
human} side of teaching interactions.  We then propose a set of
conjectures about aspects of human teaching behavior that we believe
could be incorporated into future work on HT agents.},
wwwnote={<a href="http://www.aaai.org/Symposia/Spring/sss09.php">AAAI Spring 2009 Symposium: Agents that Learn from Human Teachers</a>},
}


Generated by bib2html.pl (written by Patrick Riley ) on Mon Jul 18, 2016 17:12:24