miun.sePublikationer
Ändra sökning
RefereraExporteraLänk till posten
Permanent länk

Direktlänk
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
A Fast Bilinear Optimization Algorithm
Mittuniversitetet, Fakulteten för naturvetenskap, teknik och medier, Institutionen för informationsteknologi och medier. (FSCN - Fibre Science and Communication Network)
Mittuniversitetet, Fakulteten för naturvetenskap, teknik och medier, Institutionen för informationsteknologi och medier.ORCID-id: 0000-0002-0665-1889
2004 (Engelska)Ingår i: Proceedings of the Third International Conference on Nonlinear Analysis and Convex Analysis 2003 (NACA 2003), Yokohama: Yokohama Publishers , 2004Konferensbidrag, Publicerat paper (Refereegranskat)
Abstract [en]

Computational decision analysis methods (such as DELTA) have been developed and implemented over a number of years for solving decision problems when vague and numerically imprecise information prevails. However, the evaluation phases in those methods often give rise to bilinear programming problems, which are time-consuming to solve in an interactive environment with general nonlinear programming solvers such as SNOPT. This paper proposes a linear programming based algorithm for solving this type of problems, thus enabling interactive use of such methods.

Ort, förlag, år, upplaga, sidor
Yokohama: Yokohama Publishers , 2004.
Nyckelord [en]
linear program, bilinear program, interval probability, utility theory, decision analysis
Nationell ämneskategori
Datavetenskap (datalogi)
Identifikatorer
URN: urn:nbn:se:miun:diva-1707Lokalt ID: 356ISBN: 4-946552-15-4 (tryckt)OAI: oai:DiVA.org:miun-1707DiVA, id: diva2:26739
Tillgänglig från: 2008-09-30 Skapad: 2008-09-30 Senast uppdaterad: 2018-01-12Bibliografiskt granskad
Ingår i avhandling
1. Bilinear optimization in computational decision analysis
Öppna denna publikation i ny flik eller fönster >>Bilinear optimization in computational decision analysis
2005 (Engelska)Doktorsavhandling, sammanläggning (Övrigt vetenskapligt)
Abstract [en]

In real-life decision analysis, significant recognition has been given to theunrealistic expectation of numerically precise information. Many modernapproaches attempting to handle imprecision have focused more on representationand less on evaluation. The DELTA method, as one of the fewleading approaches, challenges this issue by its evaluation framework thatcan accommodate both precision and imprecision. However, computationally,DELTA and similar approaches may incur time-consuming calculationsdue to the introduction of imprecise information concerning probability andutility. In general, those problems are bounded non-convex bilinear optimizationprograms with disjoint linear constraints, which cannot be solvedeffectively by any existing general-purpose global optimization technique.This thesis presents two enhanced cutting plane algorithms for solvingbounded disjoint bilinear programs arising in computational decision analysis.Each algorithm consists of a local phase designed to determine a localoptimizer from an approximate solution, and a global phase designed to systematicallyexplore the feasible region, subset by subset. These two phasesare switched automatically during the global search procedure. The basicframework builds upon previously developed efficient cutting plane methods.By embedding the lower bounding technique in a branch and bound procedure,the improvement of their performances seems encouraging in the lightof computational experience. Even though the motivation to develop thesealgorithms stems from computational decision analysis, the idea can also beextended to the development of optimization approaches for handling generalbounded disjoint bilinear programs, especially for larger sized ones.

Ort, förlag, år, upplaga, sidor
Sundsvall: Mittuniversitetet, 2005
Serie
Mid Sweden University doctoral thesis, ISSN 1652-893X ; 2
Nationell ämneskategori
Systemvetenskap, informationssystem och informatik
Identifikatorer
urn:nbn:se:miun:diva-8856 (URN)
Disputation
2005-05-27, O102, Åkroken, Sundsvall, 13:15 (Engelska)
Opponent
Handledare
Tillgänglig från: 2009-05-06 Skapad: 2009-05-06 Senast uppdaterad: 2018-01-13Bibliografiskt granskad

Open Access i DiVA

Fulltext saknas i DiVA

Personposter BETA

Ding, XiaosongEkenberg, Love

Sök vidare i DiVA

Av författaren/redaktören
Ding, XiaosongEkenberg, Love
Av organisationen
Institutionen för informationsteknologi och medier
Datavetenskap (datalogi)

Sök vidare utanför DiVA

GoogleGoogle Scholar

isbn
urn-nbn

Altmetricpoäng

isbn
urn-nbn
Totalt: 169 träffar
RefereraExporteraLänk till posten
Permanent länk

Direktlänk
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf