Probabilistic planning via determinization in hindsight

Sungwook Yoon, Alan Fern, Robert Givan, Subbarao Kambhampati

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

60 Citations (Scopus)

Abstract

This paper investigates hindsight optimization as an approach for leveraging the significant advances in deterministic planning for action selection in probabilistic domains. Hindsight optimization is an online technique that evaluates the one-step-reachable states by sampling future outcomes to generate multiple non-stationary deterministic planning problems which can then be solved using search. Hindsight optimization has been successfully used in a number of online scheduling applications; however, it has not yet been considered in the substantially different context of goal-based probabilistic planning. We describe an implementation of hindsight optimization for probabilistic planning based on deterministic forward heuristic search and evaluate its performance on planning-competition benchmarks and other probabilistically interesting problems. The planner is able to outperform a number of probabilistic planners including FF-Replan on many problems. Finally, we investigate conditions under which hindsight optimization is guaranteed to be effective with respect to goal achievement, and also illustrate examples where the approach can go wrong.

Original languageEnglish (US)
Title of host publicationProceedings of the National Conference on Artificial Intelligence
Pages1010-1016
Number of pages7
Volume2
StatePublished - 2008
Event23rd AAAI Conference on Artificial Intelligence and the 20th Innovative Applications of Artificial Intelligence Conference, AAAI-08/IAAI-08 - Chicago, IL, United States
Duration: Jul 13 2008Jul 17 2008

Other

Other23rd AAAI Conference on Artificial Intelligence and the 20th Innovative Applications of Artificial Intelligence Conference, AAAI-08/IAAI-08
CountryUnited States
CityChicago, IL
Period7/13/087/17/08

Fingerprint

Planning
Scheduling
Sampling

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence

Cite this

Yoon, S., Fern, A., Givan, R., & Kambhampati, S. (2008). Probabilistic planning via determinization in hindsight. In Proceedings of the National Conference on Artificial Intelligence (Vol. 2, pp. 1010-1016)

Probabilistic planning via determinization in hindsight. / Yoon, Sungwook; Fern, Alan; Givan, Robert; Kambhampati, Subbarao.

Proceedings of the National Conference on Artificial Intelligence. Vol. 2 2008. p. 1010-1016.

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

Yoon, S, Fern, A, Givan, R & Kambhampati, S 2008, Probabilistic planning via determinization in hindsight. in Proceedings of the National Conference on Artificial Intelligence. vol. 2, pp. 1010-1016, 23rd AAAI Conference on Artificial Intelligence and the 20th Innovative Applications of Artificial Intelligence Conference, AAAI-08/IAAI-08, Chicago, IL, United States, 7/13/08.
Yoon S, Fern A, Givan R, Kambhampati S. Probabilistic planning via determinization in hindsight. In Proceedings of the National Conference on Artificial Intelligence. Vol. 2. 2008. p. 1010-1016
Yoon, Sungwook ; Fern, Alan ; Givan, Robert ; Kambhampati, Subbarao. / Probabilistic planning via determinization in hindsight. Proceedings of the National Conference on Artificial Intelligence. Vol. 2 2008. pp. 1010-1016
@inproceedings{1212de3587a349cc8d690d6922ddb273,
title = "Probabilistic planning via determinization in hindsight",
abstract = "This paper investigates hindsight optimization as an approach for leveraging the significant advances in deterministic planning for action selection in probabilistic domains. Hindsight optimization is an online technique that evaluates the one-step-reachable states by sampling future outcomes to generate multiple non-stationary deterministic planning problems which can then be solved using search. Hindsight optimization has been successfully used in a number of online scheduling applications; however, it has not yet been considered in the substantially different context of goal-based probabilistic planning. We describe an implementation of hindsight optimization for probabilistic planning based on deterministic forward heuristic search and evaluate its performance on planning-competition benchmarks and other probabilistically interesting problems. The planner is able to outperform a number of probabilistic planners including FF-Replan on many problems. Finally, we investigate conditions under which hindsight optimization is guaranteed to be effective with respect to goal achievement, and also illustrate examples where the approach can go wrong.",
author = "Sungwook Yoon and Alan Fern and Robert Givan and Subbarao Kambhampati",
year = "2008",
language = "English (US)",
isbn = "9781577353683",
volume = "2",
pages = "1010--1016",
booktitle = "Proceedings of the National Conference on Artificial Intelligence",

}

TY - GEN

T1 - Probabilistic planning via determinization in hindsight

AU - Yoon, Sungwook

AU - Fern, Alan

AU - Givan, Robert

AU - Kambhampati, Subbarao

PY - 2008

Y1 - 2008

N2 - This paper investigates hindsight optimization as an approach for leveraging the significant advances in deterministic planning for action selection in probabilistic domains. Hindsight optimization is an online technique that evaluates the one-step-reachable states by sampling future outcomes to generate multiple non-stationary deterministic planning problems which can then be solved using search. Hindsight optimization has been successfully used in a number of online scheduling applications; however, it has not yet been considered in the substantially different context of goal-based probabilistic planning. We describe an implementation of hindsight optimization for probabilistic planning based on deterministic forward heuristic search and evaluate its performance on planning-competition benchmarks and other probabilistically interesting problems. The planner is able to outperform a number of probabilistic planners including FF-Replan on many problems. Finally, we investigate conditions under which hindsight optimization is guaranteed to be effective with respect to goal achievement, and also illustrate examples where the approach can go wrong.

AB - This paper investigates hindsight optimization as an approach for leveraging the significant advances in deterministic planning for action selection in probabilistic domains. Hindsight optimization is an online technique that evaluates the one-step-reachable states by sampling future outcomes to generate multiple non-stationary deterministic planning problems which can then be solved using search. Hindsight optimization has been successfully used in a number of online scheduling applications; however, it has not yet been considered in the substantially different context of goal-based probabilistic planning. We describe an implementation of hindsight optimization for probabilistic planning based on deterministic forward heuristic search and evaluate its performance on planning-competition benchmarks and other probabilistically interesting problems. The planner is able to outperform a number of probabilistic planners including FF-Replan on many problems. Finally, we investigate conditions under which hindsight optimization is guaranteed to be effective with respect to goal achievement, and also illustrate examples where the approach can go wrong.

UR - http://www.scopus.com/inward/record.url?scp=57749193939&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=57749193939&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:57749193939

SN - 9781577353683

VL - 2

SP - 1010

EP - 1016

BT - Proceedings of the National Conference on Artificial Intelligence

ER -