We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Optimization and Control

Title: Exact polynomial optimization strengthened with Fritz John conditions

Abstract: Let $f,g_1,\dots,g_m$ be polynomials with real coefficients in a vector of variables $x=(x_1,\dots,x_n)$. Denote by $\text{diag}(g)$ the diagonal matrix with coefficients $g=(g_1,\dots,g_m)$ and denote by $\nabla g$ the Jacobian of $g$. Let $C$ be the set of critical points defined by \begin{equation}
C=\{x\in\mathbb R^n\,:\,\text{rank}(\varphi(x))< m\}\quad\text{with}\quad\varphi:=\begin{bmatrix} \nabla g\\ \text{diag}(g) \end{bmatrix}\,. \end{equation} Assume that the image of $C$ under $f$, denoted by $f(C)$, is empty or finite. (Our assumption holds generically since $C$ is empty in a Zariski open set in the space of the coefficients of $g_1,\dots,g_m$ with given degrees.) We provide a sequence of values, returned by semidefinite programs, finitely converges to the minimal value attained by $f$ over the basic semi-algebraic set $S$ defined by \begin{equation}
S:=\{x\in\mathbb R^n\,:\,g_j(x)\ge 0\,,\,j=1,\dots,m\}\,. \end{equation} Consequently, we can compute exactly the minimal value of any polynomial with real coefficients in $x$ over one of the following sets: the unit ball, the unit hypercube and the unit simplex.
Comments: 20 pages and 1 table
Subjects: Optimization and Control (math.OC); Algebraic Geometry (math.AG)
Cite as: arXiv:2205.04254 [math.OC]
  (or arXiv:2205.04254v1 [math.OC] for this version)

Submission history

From: Ngoc Hoang Anh Mai [view email]
[v1] Mon, 9 May 2022 13:13:29 GMT (20kb)
[v2] Sun, 26 Jun 2022 09:47:55 GMT (25kb)

Link back to: arXiv, form interface, contact.