Approximated perspective relaxations: a project and lift approach
Frangioni, Antonio; Furini, Fabio; Gentile, Claudio (2016), Approximated perspective relaxations: a project and lift approach, Computational Optimization and Applications, 63, 3, p. 705-735. 10.1007/s10589-015-9787-8
Type
Article accepté pour publication ou publiéDate
2016Journal name
Computational Optimization and ApplicationsVolume
63Number
3Publisher
Kluwer Academic Publishers
Pages
705-735
Publication identifier
Metadata
Show full item recordAuthor(s)
Frangioni, AntonioDepartment of Computer Science [Pisa]
Furini, Fabio
Laboratoire d'analyse et modélisation de systèmes pour l'aide à la décision [LAMSADE]
Gentile, Claudio
Istituto di Analisi dei Sistemi ed Informatica "A. Ruberti"
Abstract (EN)
The perspective reformulation (PR) of a Mixed-Integer NonLinear Program with semi-continuous variables is obtained by replacing each term in the (separable) objective function with its convex envelope. Solving the corresponding continuous relaxation requires appropriate techniques. Under some rather restrictive assumptions, the Projected PR (P2R) can be defined where the integer variables are eliminated by projecting the solution set onto the space of the continuous variables only. This approach produces a simple piecewise-convex problem with the same structure as the original one; however, this prevents the use of general-purpose solvers, in that some variables are then only implicitly represented in the formulation. We show how to construct an Approximated Projected PR (AP2R) whereby the projected formulation is “lifted” back to the original variable space, with each integer variable expressing one piece of the obtained piecewise-convex function. In some cases, this produces a reformulation of the original problem with exactly the same size and structure as the standard continuous relaxation, but providing substantially improved bounds. In the process we also substantially extend the approach beyond the original P2R development by relaxing the requirement that the objective function be quadratic and the left endpoint of the domain of the variables be non-negative. While the AP2R bound can be weaker than that of the PR, this approach can be applied in many more cases and allows direct use of off-the-shelf MINLP software; this is shown to be competitive with previously proposed approaches in some applications.Subjects / Keywords
Mixed-integer nonlinear problems; Semi-continuous variables; Perspective reformulation; ProjectionRelated items
Showing items related by title and author.
-
Frangioni, Antonio; Furini, Fabio; Gentile, Claudio (2017) Article accepté pour publication ou publié
-
Furini, Fabio; Traversi, Emiliano; Belotti, Pietro; Frangioni, Antonio; Gleixner, Ambros; Gould, Nick; Liberti, Leo; Lodi, Andrea; Misener, Ruth; Mittelmann, Hans; Sahinidis, Nikolaos V.; Vigerske, Stefan; Wiegele, Angelika (2019) Article accepté pour publication ou publié
-
Morandi, Fabio; Scaruffi, Paola; Gallo, Fabio; Stigliani, Sara; Moretti, Stefano; Bonassi, Stefano; Gambini, Claudio; Mazzocco, Katia; Fardin, Paolo; Haupt, Riccardo; Pistoia, Vito; Tonini, Gian Paolo; Corrias, Maria Valeria (2012) Article accepté pour publication ou publié
-
Fiorini, Laura; D’Onofrio, Grazia; Limosani, Raffaele; Sancarlo, Daniele; Greco, Antonio; Giuliani, Francesco; Kung, Antonio; Dario, Paolo; Cavallo, Filippo Communication / Conférence
-
Agile Co-Creation for Robots and Aging (ACCRA) Project: new technological solutions for older people D’Onofrio, Grazia; Fiorini, Laura; De Mul, Marleen; Guiot, Denis; Sengès, Eloïse; Fabbricotti, Isabelle; Okabe, Yasuo; Hoshino, Hiroshi; Limosani, Raffaele; Vitanza, Alessandra; Greco, Francesca; Giuliani, Francesco; Kung, Antonio; Cavallo, Filippo; Sancarlo, Daniele; Greco, Antonio (2018) Article accepté pour publication ou publié