Learning qualitative models by demonstration

Thomas R Hinrichs*, Kenneth D Forbus

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

Creating software agents that learn interactively requires the ability to learn from a small number of trials, extracting general, flexible knowledge that can drive behavior from observation and interaction. We claim that qualitative models provide a useful intermediate level of causal representation for dynamic domains, including the formulation of strategies and tactics. We argue that qualitative models are quickly learnable, and enable model based reasoning techniques to be used to recognize, operationalize, and construct more strategic knowledge. This paper describes an approach to incrementally learning qualitative influences by demonstration in the context of a strategy game. We show how the learned model can help a system play by enabling it to explain which actions could contribute to maximizing a quantitative goal. We also show how reasoning about the model allows it to reformulate a learning problem to address delayed effects and credit assignment, such that it can improve its performance on more strategic tasks such as city placement.

Original languageEnglish (US)
Title of host publicationAAAI-12 / IAAI-12 - Proceedings of the 26th AAAI Conference on Artificial Intelligence and the 24th Innovative Applications of Artificial Intelligence Conference
Pages207-213
Number of pages7
Volume1
StatePublished - Nov 7 2012
Event26th AAAI Conference on Artificial Intelligence and the 24th Innovative Applications of Artificial Intelligence Conference, AAAI-12 / IAAI-12 - Toronto, ON, Canada
Duration: Jul 22 2012Jul 26 2012

Other

Other26th AAAI Conference on Artificial Intelligence and the 24th Innovative Applications of Artificial Intelligence Conference, AAAI-12 / IAAI-12
CountryCanada
CityToronto, ON
Period7/22/127/26/12

Fingerprint

Demonstrations
Software agents

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence

Cite this

Hinrichs, T. R., & Forbus, K. D. (2012). Learning qualitative models by demonstration. In AAAI-12 / IAAI-12 - Proceedings of the 26th AAAI Conference on Artificial Intelligence and the 24th Innovative Applications of Artificial Intelligence Conference (Vol. 1, pp. 207-213)
Hinrichs, Thomas R ; Forbus, Kenneth D. / Learning qualitative models by demonstration. AAAI-12 / IAAI-12 - Proceedings of the 26th AAAI Conference on Artificial Intelligence and the 24th Innovative Applications of Artificial Intelligence Conference. Vol. 1 2012. pp. 207-213
@inproceedings{cc18c99c8c0b4439914370aeec7d4205,
title = "Learning qualitative models by demonstration",
abstract = "Creating software agents that learn interactively requires the ability to learn from a small number of trials, extracting general, flexible knowledge that can drive behavior from observation and interaction. We claim that qualitative models provide a useful intermediate level of causal representation for dynamic domains, including the formulation of strategies and tactics. We argue that qualitative models are quickly learnable, and enable model based reasoning techniques to be used to recognize, operationalize, and construct more strategic knowledge. This paper describes an approach to incrementally learning qualitative influences by demonstration in the context of a strategy game. We show how the learned model can help a system play by enabling it to explain which actions could contribute to maximizing a quantitative goal. We also show how reasoning about the model allows it to reformulate a learning problem to address delayed effects and credit assignment, such that it can improve its performance on more strategic tasks such as city placement.",
author = "Hinrichs, {Thomas R} and Forbus, {Kenneth D}",
year = "2012",
month = "11",
day = "7",
language = "English (US)",
isbn = "9781577355687",
volume = "1",
pages = "207--213",
booktitle = "AAAI-12 / IAAI-12 - Proceedings of the 26th AAAI Conference on Artificial Intelligence and the 24th Innovative Applications of Artificial Intelligence Conference",

}

Hinrichs, TR & Forbus, KD 2012, Learning qualitative models by demonstration. in AAAI-12 / IAAI-12 - Proceedings of the 26th AAAI Conference on Artificial Intelligence and the 24th Innovative Applications of Artificial Intelligence Conference. vol. 1, pp. 207-213, 26th AAAI Conference on Artificial Intelligence and the 24th Innovative Applications of Artificial Intelligence Conference, AAAI-12 / IAAI-12, Toronto, ON, Canada, 7/22/12.

Learning qualitative models by demonstration. / Hinrichs, Thomas R; Forbus, Kenneth D.

AAAI-12 / IAAI-12 - Proceedings of the 26th AAAI Conference on Artificial Intelligence and the 24th Innovative Applications of Artificial Intelligence Conference. Vol. 1 2012. p. 207-213.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Learning qualitative models by demonstration

AU - Hinrichs, Thomas R

AU - Forbus, Kenneth D

PY - 2012/11/7

Y1 - 2012/11/7

N2 - Creating software agents that learn interactively requires the ability to learn from a small number of trials, extracting general, flexible knowledge that can drive behavior from observation and interaction. We claim that qualitative models provide a useful intermediate level of causal representation for dynamic domains, including the formulation of strategies and tactics. We argue that qualitative models are quickly learnable, and enable model based reasoning techniques to be used to recognize, operationalize, and construct more strategic knowledge. This paper describes an approach to incrementally learning qualitative influences by demonstration in the context of a strategy game. We show how the learned model can help a system play by enabling it to explain which actions could contribute to maximizing a quantitative goal. We also show how reasoning about the model allows it to reformulate a learning problem to address delayed effects and credit assignment, such that it can improve its performance on more strategic tasks such as city placement.

AB - Creating software agents that learn interactively requires the ability to learn from a small number of trials, extracting general, flexible knowledge that can drive behavior from observation and interaction. We claim that qualitative models provide a useful intermediate level of causal representation for dynamic domains, including the formulation of strategies and tactics. We argue that qualitative models are quickly learnable, and enable model based reasoning techniques to be used to recognize, operationalize, and construct more strategic knowledge. This paper describes an approach to incrementally learning qualitative influences by demonstration in the context of a strategy game. We show how the learned model can help a system play by enabling it to explain which actions could contribute to maximizing a quantitative goal. We also show how reasoning about the model allows it to reformulate a learning problem to address delayed effects and credit assignment, such that it can improve its performance on more strategic tasks such as city placement.

UR - http://www.scopus.com/inward/record.url?scp=84868288435&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84868288435&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9781577355687

VL - 1

SP - 207

EP - 213

BT - AAAI-12 / IAAI-12 - Proceedings of the 26th AAAI Conference on Artificial Intelligence and the 24th Innovative Applications of Artificial Intelligence Conference

ER -

Hinrichs TR, Forbus KD. Learning qualitative models by demonstration. In AAAI-12 / IAAI-12 - Proceedings of the 26th AAAI Conference on Artificial Intelligence and the 24th Innovative Applications of Artificial Intelligence Conference. Vol. 1. 2012. p. 207-213