On variable dependencies and compressed pattern databases

Malte Helmert; Nathan R. Sturtevant; Ariel Felner

doi:10.1609/socs.v8i1.18437

On variable dependencies and compressed pattern databases

Malte Helmert, Nathan R. Sturtevant, Ariel Felner

Research output: Contribution to conference › Paper › peer-review

1 Scopus citations

Abstract

Pattern databases are among the strongest known heuristics for many classical search benchmarks such as sliding-tile puzzles, the 4-peg Towers of Hanoi puzzles, Rubik’s Cube, and TopSpin. Min-compression is a generally applicable technique for augmenting pattern database heuristics that has led to marked experimental improvements in some settings, while being ineffective in others. We provide a theoretical explanation for these experimental phenomena by studying the interaction between the ranking function used to order abstract states in a pattern database, the compression scheme used to abstract states, and the dependencies between state variables in the problem representation.

Original language	English
Pages	129-133
Number of pages	5
DOIs	https://doi.org/10.1609/socs.v8i1.18437
State	Published - 2017
Externally published	Yes
Event	10th Annual Symposium on Combinatorial Search, SoCS 2017 - Pittsburgh, United States Duration: 16 Jun 2017 → 17 Jun 2017

Conference

Conference	10th Annual Symposium on Combinatorial Search, SoCS 2017
Country/Territory	United States
City	Pittsburgh
Period	16/06/17 → 17/06/17

Bibliographical note

Publisher Copyright:
Copyright c 2017, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.

Funding

We thank the anonymous reviewers for their helpful comments. This work was supported by the Swiss National Science Foundation (SNSF) as part of the project “Reasoning about Plans and Heuristics for Planning and Combinatorial Search” (RAPAHPACS), by the National Science Foundation (NSF) under Grant No. 1551406 and by Israel Science Foundation (ISF) grant #417/13.

Funders	Funder number
National Science Foundation	1551406
Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung
Israel Science Foundation	417/13

Access to Document

10.1609/socs.v8i1.18437

Cite this

@conference{33365a48944e4c8aa294e86901d1f3d0,

title = "On variable dependencies and compressed pattern databases",

abstract = "Pattern databases are among the strongest known heuristics for many classical search benchmarks such as sliding-tile puzzles, the 4-peg Towers of Hanoi puzzles, Rubik{\textquoteright}s Cube, and TopSpin. Min-compression is a generally applicable technique for augmenting pattern database heuristics that has led to marked experimental improvements in some settings, while being ineffective in others. We provide a theoretical explanation for these experimental phenomena by studying the interaction between the ranking function used to order abstract states in a pattern database, the compression scheme used to abstract states, and the dependencies between state variables in the problem representation.",

author = "Malte Helmert and Sturtevant, {Nathan R.} and Ariel Felner",

note = "Publisher Copyright: Copyright c 2017, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.; 10th Annual Symposium on Combinatorial Search, SoCS 2017 ; Conference date: 16-06-2017 Through 17-06-2017",

year = "2017",

doi = "10.1609/socs.v8i1.18437",

language = "אנגלית",

pages = "129--133",

}

TY - CONF

T1 - On variable dependencies and compressed pattern databases

AU - Helmert, Malte

AU - Sturtevant, Nathan R.

AU - Felner, Ariel

PY - 2017

Y1 - 2017

N2 - Pattern databases are among the strongest known heuristics for many classical search benchmarks such as sliding-tile puzzles, the 4-peg Towers of Hanoi puzzles, Rubik’s Cube, and TopSpin. Min-compression is a generally applicable technique for augmenting pattern database heuristics that has led to marked experimental improvements in some settings, while being ineffective in others. We provide a theoretical explanation for these experimental phenomena by studying the interaction between the ranking function used to order abstract states in a pattern database, the compression scheme used to abstract states, and the dependencies between state variables in the problem representation.

AB - Pattern databases are among the strongest known heuristics for many classical search benchmarks such as sliding-tile puzzles, the 4-peg Towers of Hanoi puzzles, Rubik’s Cube, and TopSpin. Min-compression is a generally applicable technique for augmenting pattern database heuristics that has led to marked experimental improvements in some settings, while being ineffective in others. We provide a theoretical explanation for these experimental phenomena by studying the interaction between the ranking function used to order abstract states in a pattern database, the compression scheme used to abstract states, and the dependencies between state variables in the problem representation.

UR - http://www.scopus.com/inward/record.url?scp=85050528146&partnerID=8YFLogxK

U2 - 10.1609/socs.v8i1.18437

DO - 10.1609/socs.v8i1.18437

M3 - ???researchoutput.researchoutputtypes.contributiontoconference.paper???

AN - SCOPUS:85050528146

SP - 129

EP - 133

T2 - 10th Annual Symposium on Combinatorial Search, SoCS 2017

Y2 - 16 June 2017 through 17 June 2017

ER -

On variable dependencies and compressed pattern databases

Abstract

Conference

Bibliographical note

Funding

Access to Document

Other files and links

Fingerprint

Cite this