Testing the non-diagonal quadratic convex reformulation technique

Otto Nissfolk; Ray Pörn; Tapio Westerlund

doi:10.1016/B978-0-444-63428-3.50060-6

Testing the non-diagonal quadratic convex reformulation technique

Otto Nissfolk, Ray Pörn, Tapio Westerlund

Research output: Chapter in Book/Conference proceeding › Conference contribution › Scientific › peer-review

2 Citations (Scopus)

Abstract

Ji et al. (2012) introduced a new reformulation technique for general 0-1 quadratic programs. They did not name it so we call it Non-Diagonal Quadratic Convex Reformulation (NDQCR). The reformulation technique is based on the Quadratic Convex Reformulation method developed by Billionnet et al. (2009, 2012, 2013). In this paper we test the NDQCR method. Specifically we test how the number of included non-diagonal elements affect the solution times for solved problems and also the solution qualities for problems not solved within the time-limit. We also present a new best known lower bound for the largest problem in the QAPLIB (2013), the tai256c problem introduced by Taillard (1995).

Original language	Undefined/Unknown
Title of host publication	26th European Symposium on Computer Aided Process Engineering
Editors	Zdravko Kravanja
Publisher	Elsevier
Pages	331–336
ISBN (Electronic)	9780444634443
ISBN (Print)	9780444634283
DOIs	https://doi.org/10.1016/B978-0-444-63428-3.50060-6
Publication status	Published - 2016
MoE publication type	A4 Article in a conference publication
Event	European Symposium on Computer Aided Process Engineering - 26th European Symposium on Computer Aided Process Engineering Duration: 12 Jun 2016 → 15 Jun 2016

Conference

Conference	European Symposium on Computer Aided Process Engineering
Period	12/06/16 → 15/06/16

Access to Document

10.1016/B978-0-444-63428-3.50060-6

http://www.sciencedirect.com/science/article/pii/B9780444634283500606

Cite this

@inproceedings{6cecab9881f04627a21f87ee6779841d,

title = "Testing the non-diagonal quadratic convex reformulation technique",

abstract = "Ji et al. (2012) introduced a new reformulation technique for general 0-1 quadratic programs. They did not name it so we call it Non-Diagonal Quadratic Convex Reformulation (NDQCR). The reformulation technique is based on the Quadratic Convex Reformulation method developed by Billionnet et al. (2009, 2012, 2013). In this paper we test the NDQCR method. Specifically we test how the number of included non-diagonal elements affect the solution times for solved problems and also the solution qualities for problems not solved within the time-limit. We also present a new best known lower bound for the largest problem in the QAPLIB (2013), the tai256c problem introduced by Taillard (1995).",

author = "Otto Nissfolk and Ray P{\"o}rn and Tapio Westerlund",

note = "ast. 26th European Symposium on Computer Aided Process Engineering contains the papers presented at the 26th European Society of Computer-Aided Process Engineering (ESCAPE) Event held at Portoroz Slovenia, from June 12th to June 15th, 2016.; European Symposium on Computer Aided Process Engineering ; Conference date: 12-06-2016 Through 15-06-2016",

year = "2016",

doi = "10.1016/B978-0-444-63428-3.50060-6",

language = "Odefinierat/ok{\"a}nt",

isbn = "9780444634283",

pages = "331–336",

editor = "Zdravko Kravanja",

booktitle = "26th European Symposium on Computer Aided Process Engineering",

publisher = "Elsevier",

}

TY - GEN

T1 - Testing the non-diagonal quadratic convex reformulation technique

AU - Nissfolk, Otto

AU - Pörn, Ray

AU - Westerlund, Tapio

N1 - ast. 26th European Symposium on Computer Aided Process Engineering contains the papers presented at the 26th European Society of Computer-Aided Process Engineering (ESCAPE) Event held at Portoroz Slovenia, from June 12th to June 15th, 2016.

PY - 2016

Y1 - 2016

N2 - Ji et al. (2012) introduced a new reformulation technique for general 0-1 quadratic programs. They did not name it so we call it Non-Diagonal Quadratic Convex Reformulation (NDQCR). The reformulation technique is based on the Quadratic Convex Reformulation method developed by Billionnet et al. (2009, 2012, 2013). In this paper we test the NDQCR method. Specifically we test how the number of included non-diagonal elements affect the solution times for solved problems and also the solution qualities for problems not solved within the time-limit. We also present a new best known lower bound for the largest problem in the QAPLIB (2013), the tai256c problem introduced by Taillard (1995).

AB - Ji et al. (2012) introduced a new reformulation technique for general 0-1 quadratic programs. They did not name it so we call it Non-Diagonal Quadratic Convex Reformulation (NDQCR). The reformulation technique is based on the Quadratic Convex Reformulation method developed by Billionnet et al. (2009, 2012, 2013). In this paper we test the NDQCR method. Specifically we test how the number of included non-diagonal elements affect the solution times for solved problems and also the solution qualities for problems not solved within the time-limit. We also present a new best known lower bound for the largest problem in the QAPLIB (2013), the tai256c problem introduced by Taillard (1995).

U2 - 10.1016/B978-0-444-63428-3.50060-6

DO - 10.1016/B978-0-444-63428-3.50060-6

M3 - Konferensbidrag

SN - 9780444634283

SP - 331

EP - 336

BT - 26th European Symposium on Computer Aided Process Engineering

A2 - Kravanja, Zdravko

PB - Elsevier

T2 - European Symposium on Computer Aided Process Engineering

Y2 - 12 June 2016 through 15 June 2016

ER -