Why adding more constraints makes a problem easier for hill-climbing algorithms

Analyzing landscapes of CSPs

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

37 Citations (Scopus)

Abstract

It is well known that constraint satisfaction problems (CSPs) in the phase transition region are most difficult for complete search algorithms. On the other hand, for incomplete hill-climbing algorithms, problems in the phase transition region axe more difficult than problems beyond the phase transition region, i.e., more constrained problems. This result seems somewhat unnatural since these more constrained problems have fewer solutions than the phase transition problems. In this paper, we clarify the cause of this paradoxical phenomenon by exhaustively analyzing the state-space landscape of CSPs, which is formed by the evaluation values of states. The analyses show that by adding more constraints, while the number of solutions decreases, the number of local-minima also decreases, thus the number of states that are reachable to solutions increases. Furthermore, the analyses clarify that the decrease in local-minima is caused by a set of interconnected local-minima (basin) being divided into smaller regions by adding more constraints.

Original languageEnglish
Title of host publicationPrinciples and Practice of Constraint Programming - CP 1997 - 3rd International Conference, CP 1997, Proceedings
PublisherSpringer Verlag
Pages356-370
Number of pages15
Volume1330
ISBN (Print)3540637532, 9783540637530
Publication statusPublished - 1997
Externally publishedYes
Event3rd International Conference on Principles and Practice of Constraint Programming, CP 1997 - Linz, Austria
Duration: Oct 29 1997Nov 1 1997

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1330
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other3rd International Conference on Principles and Practice of Constraint Programming, CP 1997
CountryAustria
CityLinz
Period10/29/9711/1/97

Fingerprint

Constraint satisfaction problems
Hill Climbing
Constraint Satisfaction Problem
Phase transitions
Phase Transition
Local Minima
Decrease
Number of Solutions
Search Algorithm
State Space
Local search (optimization)
Evaluation

All Science Journal Classification (ASJC) codes

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Yokoo, M. (1997). Why adding more constraints makes a problem easier for hill-climbing algorithms: Analyzing landscapes of CSPs. In Principles and Practice of Constraint Programming - CP 1997 - 3rd International Conference, CP 1997, Proceedings (Vol. 1330, pp. 356-370). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1330). Springer Verlag.

Why adding more constraints makes a problem easier for hill-climbing algorithms : Analyzing landscapes of CSPs. / Yokoo, Makoto.

Principles and Practice of Constraint Programming - CP 1997 - 3rd International Conference, CP 1997, Proceedings. Vol. 1330 Springer Verlag, 1997. p. 356-370 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1330).

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

Yokoo, M 1997, Why adding more constraints makes a problem easier for hill-climbing algorithms: Analyzing landscapes of CSPs. in Principles and Practice of Constraint Programming - CP 1997 - 3rd International Conference, CP 1997, Proceedings. vol. 1330, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1330, Springer Verlag, pp. 356-370, 3rd International Conference on Principles and Practice of Constraint Programming, CP 1997, Linz, Austria, 10/29/97.
Yokoo M. Why adding more constraints makes a problem easier for hill-climbing algorithms: Analyzing landscapes of CSPs. In Principles and Practice of Constraint Programming - CP 1997 - 3rd International Conference, CP 1997, Proceedings. Vol. 1330. Springer Verlag. 1997. p. 356-370. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Yokoo, Makoto. / Why adding more constraints makes a problem easier for hill-climbing algorithms : Analyzing landscapes of CSPs. Principles and Practice of Constraint Programming - CP 1997 - 3rd International Conference, CP 1997, Proceedings. Vol. 1330 Springer Verlag, 1997. pp. 356-370 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{3c5de48d16c34c39b8931982b9340fd2,
title = "Why adding more constraints makes a problem easier for hill-climbing algorithms: Analyzing landscapes of CSPs",
abstract = "It is well known that constraint satisfaction problems (CSPs) in the phase transition region are most difficult for complete search algorithms. On the other hand, for incomplete hill-climbing algorithms, problems in the phase transition region axe more difficult than problems beyond the phase transition region, i.e., more constrained problems. This result seems somewhat unnatural since these more constrained problems have fewer solutions than the phase transition problems. In this paper, we clarify the cause of this paradoxical phenomenon by exhaustively analyzing the state-space landscape of CSPs, which is formed by the evaluation values of states. The analyses show that by adding more constraints, while the number of solutions decreases, the number of local-minima also decreases, thus the number of states that are reachable to solutions increases. Furthermore, the analyses clarify that the decrease in local-minima is caused by a set of interconnected local-minima (basin) being divided into smaller regions by adding more constraints.",
author = "Makoto Yokoo",
year = "1997",
language = "English",
isbn = "3540637532",
volume = "1330",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "356--370",
booktitle = "Principles and Practice of Constraint Programming - CP 1997 - 3rd International Conference, CP 1997, Proceedings",
address = "Germany",

}

TY - GEN

T1 - Why adding more constraints makes a problem easier for hill-climbing algorithms

T2 - Analyzing landscapes of CSPs

AU - Yokoo, Makoto

PY - 1997

Y1 - 1997

N2 - It is well known that constraint satisfaction problems (CSPs) in the phase transition region are most difficult for complete search algorithms. On the other hand, for incomplete hill-climbing algorithms, problems in the phase transition region axe more difficult than problems beyond the phase transition region, i.e., more constrained problems. This result seems somewhat unnatural since these more constrained problems have fewer solutions than the phase transition problems. In this paper, we clarify the cause of this paradoxical phenomenon by exhaustively analyzing the state-space landscape of CSPs, which is formed by the evaluation values of states. The analyses show that by adding more constraints, while the number of solutions decreases, the number of local-minima also decreases, thus the number of states that are reachable to solutions increases. Furthermore, the analyses clarify that the decrease in local-minima is caused by a set of interconnected local-minima (basin) being divided into smaller regions by adding more constraints.

AB - It is well known that constraint satisfaction problems (CSPs) in the phase transition region are most difficult for complete search algorithms. On the other hand, for incomplete hill-climbing algorithms, problems in the phase transition region axe more difficult than problems beyond the phase transition region, i.e., more constrained problems. This result seems somewhat unnatural since these more constrained problems have fewer solutions than the phase transition problems. In this paper, we clarify the cause of this paradoxical phenomenon by exhaustively analyzing the state-space landscape of CSPs, which is formed by the evaluation values of states. The analyses show that by adding more constraints, while the number of solutions decreases, the number of local-minima also decreases, thus the number of states that are reachable to solutions increases. Furthermore, the analyses clarify that the decrease in local-minima is caused by a set of interconnected local-minima (basin) being divided into smaller regions by adding more constraints.

UR - http://www.scopus.com/inward/record.url?scp=84948981891&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84948981891&partnerID=8YFLogxK

M3 - Conference contribution

SN - 3540637532

SN - 9783540637530

VL - 1330

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 356

EP - 370

BT - Principles and Practice of Constraint Programming - CP 1997 - 3rd International Conference, CP 1997, Proceedings

PB - Springer Verlag

ER -