English

Gaining or losing perspective

Optimization and Control 2020-01-07 v1

Abstract

We study MINLO (mixed-integer nonlinear optimization) formulations of the disjunction x{0}[l,u]x\in\{0\}\cup[l,u], where zz is a binary indicatorof x[l,u]x\in[l,u] (u>>0u> \ell > 0), and yy "captures" f(x)f(x), which is assumed to be convex on its domain [l,u][l,u], but otherwise y=0y=0 when x=0x=0. This model is useful when activities have operating ranges, we pay a fixed cost for carrying out each activity, and costs on the levels of activities are convex. Using volume as a measure to compare convex bodies, we investigate a variety of continuous relaxations of this model, one of which is the convex-hull, achieved via the "perspective reformulation" inequality yzf(x/z)y \geq zf(x/z). We compare this to various weaker relaxations, studying when they may be considered as viable alternatives. In the important special case when f(x):=xpf(x) := x^p, for p>1p>1, relaxations utilizing the inequality yzqxpyz^q \geq x^p, for q[0,p1]q \in [0,p-1], are higher-dimensional power-cone representable, and hence tractable in theory. One well-known concrete application (with f(x):=x2f(x) := x^2) is mean-variance optimization (in the style of Markowitz), and we carry out some experiments to illustrate our theory on this application.

Keywords

Cite

@article{arxiv.2001.01435,
  title  = {Gaining or losing perspective},
  author = {Jon Lee and Daphne Skipper and Emily Speakman},
  journal= {arXiv preprint arXiv:2001.01435},
  year   = {2020}
}
R2 v1 2026-06-23T13:03:36.324Z