Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ Universität Paderbor...arrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
https://dx.doi.org/10.17619/un...
Other literature type . 2017
Data sources: Datacite
versions View all 2 versions
addClaim

Minimization of Lipschitzian piecewise smooth objective functions

Authors: Fiege, Sabrina;

Minimization of Lipschitzian piecewise smooth objective functions

Abstract

Nichtglattheit ist eine typische Eigenschaft vieler Optimierungsprobleme, die ihren Ursprung sowohl in industriellen als auch in akademischen Anwendungen haben. Bekannte Beispiele sind Minimax-Probleme aus der Robusten Optimierung sowie die Umformulierung beschränkter Optimierungsprobleme in unbeschränkte Probleme indem man die Beschränkungen als nichtglatten Strafterme additiv zur Zielfunktion hinzufügt. Obwohl es viele Veröffentlichungen zur nichtglatter Analysis und Optimierung gibt, sind nur wenige Software-Pakete verfügbar. Daher ist das Ziel dieser Dissertation die Entwicklung und Implementierung eines Algorithmus zur Lösung unbeschränkter, nichtkonvexer und nichtglatter Optimierungsprobleme. Es wird angenommen, dass alle Nichtdifferenzierbarkeiten der Zielfunktion durch den Absolutbetrag verursacht werden. Dies umfasst auch Funktionen wie die Minimums- und Maximumsfunktion.Die Idee des entwickelten Optimierungsalgorithmus LiPsMin ist die Minimierung einer zusammengesetzten stückweise differenzierbaren Funktionen durch wiederholtes Generieren einer stückweisen Linearisierung. Dieses lokale Modell wird durch einen quadratische Term überschätzt. Die Minimierung des lokalen Modells profitiert von den zusätzlichen Informationen, die durch Strukturausnutzung gewonnen werden. Die Untersuchung von LiPsMin wird durch Konvergenzergebnisse bzgl. optimaler Punkte erster Ordnung abgerundet. Abschließend wird die numerische Effizienz des Algorithmus untersucht.

Nonsmoothness is a typical characteristic of numerous optimization problems originating from both real world and scientific applications. Well known examples from practical optimization are minimax problems used in robust optimization and the reformulation of a constrained optimization problem by adding nonsmooth penalty terms of constraint violations to the original function. Although there are plenty of publications dealing with nonsmooth analysis and optimization, there are only a few software tools available for nonsmooth optimization problems. Therefore, the purpose of this thesis is to develop, implement, and examine an algorithm for unconstrained, nonconvex, and nonsmooth optimization problems. It will be assumed that all nondifferentiabilities occurring in the objective function are caused by the absolute value and those functions that can be expressed in terms of the absolute value as the maximum and minimum function. The idea of the developed optimization method LiPsMin is the minimization of composite piecewise differentiable objective functions via successive piecewise linearization overestimated by a quadratic term. The minimization of the resulting local quadratic subproblem benefits from additional information obtained by exploiting the structure of the underlying piecewise linearization. Convergence results of LiPsMin towards first order optimal points are developed and the numerical performance of the algorithm is investigated.

Country
Germany
Related Organizations
  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Green