2007
DOI: 10.1287/ijoc.1050.0152
|View full text |Cite
|
Sign up to set email alerts
|

Performance Guarantees of Local Search for Multiprocessor Scheduling

Abstract: Increasing interest has recently been shown in analyzing the worst-case behavior of local search algorithms. In particular, the quality of local optima and the time needed to find the local optima by the simplest form of local search has been studied. This paper deals with worst-case performance of local search algorithms for makespan minimization on parallel machines. We analyze the quality of the local optima obtained by iterative improvement over the jump, swap, multi-exchange, and the newly defined push ne… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
47
0

Year Published

2009
2009
2017
2017

Publication Types

Select...
3
3

Relationship

2
4

Authors

Journals

citations
Cited by 67 publications
(47 citation statements)
references
References 31 publications
0
47
0
Order By: Relevance
“…PROPOSITION 2.1: (Finn and Horowitz [5]; Schuurman and Vredeveld [17]) The worst‐case performance guarantee of any jump‐optimal schedule and of any lex‐jump‐optimal schedule for P || C max is 2 ‐ 2/( m + 1) . This bound is tight for both neighborhoods.…”
Section: Preliminaries On Schedules and Neighborhoodsmentioning
confidence: 99%
See 2 more Smart Citations
“…PROPOSITION 2.1: (Finn and Horowitz [5]; Schuurman and Vredeveld [17]) The worst‐case performance guarantee of any jump‐optimal schedule and of any lex‐jump‐optimal schedule for P || C max is 2 ‐ 2/( m + 1) . This bound is tight for both neighborhoods.…”
Section: Preliminaries On Schedules and Neighborhoodsmentioning
confidence: 99%
“…A natural neighborhood for our scheduling problem P || C max is the so‐called jump neighborhood; in this jump neighborhood schedule σ ′ is a neighbor of schedule σ , if σ ′ results from σ by moving a single job from one machine to another machine. Finn and Horowitz [5] showed that a local optimum with respect to the jump neighborhood has a performance guarantee of 2 ‐ 2/( m + 1), and Schuurman and Vredeveld [17] showed that this bound is tight. The number of moves needed to find a jump‐optimal solution by means of an iterative improvement procedure is \documentclass{article} \usepackage{mathrsfs} \usepackage{amsmath, amssymb} \pagestyle{empty} \begin{document} \begin{align*}\mathcal{O}(n^2)\end{align*} \end{document} (as shown by Brucker et al [3]) and there exist instances for which this bound is asymptotically tight (as shown by Hurkens and Vredeveld [12]).…”
Section: Introductionmentioning
confidence: 99%
See 1 more Smart Citation
“…The price of anarchy for sequencing models when SPT policy is used is: -2 − 2/(m + 1) for m identical machines [19], -Θ(log m) for m related (restricted) machines [14], at most m for m unrelated machines [14].…”
Section: Related Workmentioning
confidence: 99%
“…The result in [14] is derived from the bound for the classical Ibarra-Kim algorithm which produces a locally optimum solution for the scheduling problem [19]. We like to point out that tere is a subtle but important difference between a locally optimal solution and a Nash solution.…”
Section: Corollary 11 the Price Of Stability In Scheduling Games Witmentioning
confidence: 99%