arXiv subjects

Aljaž Zalar

Publications and source records attributed to Aljaž Zalar.

31 records · Page 2Linked to original sources

arXiv2024

A random copositive matrix is completely positive with positive probability

An $n\times n$ symmetric matrix $A$ is copositive if the quadratic form $x^TAx$ is nonnegative on the nonnegative orthant. The cone of copositive matrices strictly contains the cone of completely positive matrices, i.e., all matrices of the form $BB^T$ for some (possibly rectangular) matrix $B$ with nonnegative entries. The main result, proved using Blekherman's real algebraic geometry inspired techniques and tools of convex geometry, shows that asymptotically, as $n$ goes to infinity, the ratio of volume radii of the two cones is strictly positive. Consequently, the same holds true for the ratio of volume radii of any two cones sandwiched between them, e.g., the cones of positive semidefinite matrices, matrices with nonnegative entries, their intersection and their Minkowski sum.

math.FA↗

arXiv2023

The truncated moment problem on curves $y=q(x)$ and $yx^\ell=1$

In this paper we study the bivariate truncated moment problem (TMP) on curves of the form $y=q(x)$, $q(x)\in \mathbb{R}[x]$, $\text{deg } q\geq 3$, and $yx^\ell=1$, $\ell\in \mathbb{N}\setminus\{1\}$. For even degree sequences the solution based on the number of moment matrix extensions was first given by Fialkow using the truncated Riesz-Haviland theorem and a sum-of-squares representations for polynomials, strictly positive on such curves. Namely, the upper bound on this number is quadratic in the degrees of the sequence and the polynomial determining a curve. We use a reduction to the univariate setting technique and improve Fialkow's bound to $\text{deg }q-1$ (resp. $\ell+1$) for curves $y=q(x)$ (resp. $yx^\ell=1$). This in turn gives analogous improvements of the degrees in the sum-of-squares representations referred to above. Moreover, we get the upper bounds on the number of atoms in the minimal representing measure, which are $k\;\text{deg }q$ (resp. $k(\ell+1)$) for curves $y=q(x)$ (resp. $yx^\ell=1$) for even degree sequences, while for odd ones they are $k\;\text{deg }q-\big\lceil\frac{\text{deg }q}{2} \big\rceil$ (resp. $k(\ell+1)-\big\lfloor\frac{\ell}{2} \big\rfloor+1$) for curves $y=q(x)$ (resp. $yx^\ell=1$). In the even case these are counterparts to the result by Riener and Schweighofer, which gives the same bound for odd degree sequences on all plane curves, while in the odd case it is a slight improvement of their bound in these special cases. Further on, we give another solution to the TMP on the curves studied based on the feasibility of a linear matrix inequality, corresponding to the univariate sequence obtained, and finally we solve concretely odd degree cases of the TMP on curves $y=x^\ell$, $\ell=2,3$, and add a new solvability condition to the even degree case on the curve $y=x^2$.

math.FA↗

arXiv2022

The strong truncated Hamburger moment problem with and without gaps

The strong truncated Hamburger moment problem (STHMP) of degree $(-2k_1,2k_2)$ asks to find necessary and sufficient conditions for the existence of a positive Borel measure, supported on $\mathbb{R}\setminus \{0\}$, such that $β_i=\int x^idμ\; (-2k_1\leq i\leq 2k_2)$. Using the solution of the truncated Hamburger moment problem and the properties of Hankel matrices we solve the STHMP. Then, using the equivalence with the STHMP of degree $(-2k,2k)$, we obtain the solution of the 2-dimensional truncated moment problem (TMP) of degree $2k$ with variety $xy=1$, first solved by Curto and Fialkow. Our addition to their result is the fact previously known only for $k=2$, that the existence of a measure is equivalent to the existence of a flat extension of the moment matrix. Further on, we solve the STHMP of degree $(-2k_1,2k_2)$ with one missing moment in the sequence, i.e., $β_{-2k_1+1}$ or $β_{2k_2-1}$, which also gives the solution of the TMP with variety $x^2y=1$ as a special case, first studied by Fialkow.

math.FA↗

arXiv2022

The truncated Hamburger moment problems with gaps in the index set

In this article we solve four special cases of the truncated Hamburger moment problem (THMP) of degree $2k$ with one or two missing moments in the sequence. As corollaries we obtain, by using appropriate substitutions, the solutions to bivariate truncated moment problems of degree $2k$ for special curves. Namely, for the curves $y=x^3$ (first solved by Fialkow), $y^2=x^3$, $y=x^4$ where a certain moment of degree $2k+1$ is known and $y^3=x^4$ with a certain moment given. The main technique is the completion of the partial positive semidefinite matrix (ppsd) such that the conditions of Curto and Fialkow's solution of the THMP are satisfied. The main tools are the use of the properties of positive semidefinite Hankel matrices and a result on all completions of a ppsd matrix with one unknown entry, proved by the use of the Schur complements for $2\times 2$ and $3\times 3$ block matrices.

math.FA↗

arXiv2022

The truncated moment problem on the union of parallel lines

In this article we study the bivariate truncated moment problem (TMP) of degree $2k$ on the union of parallel lines. First we present an alternative proof of Fialkow's solution \cite{Fia15} to the TMP on the union of two parallel lines (TMP--2pl) using the solution of the truncated Hamburger moment problem (THMP). We add a new equivalent solvability condition, which is then used together with the THMP, to solve the TMP on the union of three parallel lines (TMP--3pl), our second main result of the article. Finally, we establish a sufficient condition for the existence of a solution to the TMP on the union of $n$ parallel lines in the pure case, i.e.\ when the moment matrix $M_k$ is of the highest possible rank, or equivalently the only column relations come from the union of $n$ lines. The condition is based on the feasibility of a certain linear matrix inequality, corresponding to the extension of $M_k$ by adding rows and columns indexed by some monomials of degree $k+1$. The proof is by induction on $n$, where $n\geq 2$ and for the base of induction $n=2$ we use the solution of the TMP--2pl.

math.FA↗

arXiv2020

A local-global principle for linear dependence in enveloping algebras of Lie algebras

For every associative algebra $A$ and every class $\mathcal{C}$ of representations of $A$ the following question (related to nullstellensatz) makes sense: Characterize all tuples of elements $a_1,\ldots,a_n \in A$ such that vectors $π(a_1)v,\ldots,π(a_n)v$ are linearly dependent for every $π\in \mathcal{C}$ and every $v$ from the representation space of $π$. We answer this question in the following cases: (1) $A=U(L)$ is the enveloping algebra of a finite-dimensional complex Lie algebra $L$ and $\mathbb{C}$ is the class of all finite-dimensional representations of $A$. (2) $A=U(\mathfrak{sl}_2(\mathbb{C}))$ and $\mathbb{C}$ is the class of all finite-dimensional irreducible representations of $A$. (3) $A=U(\mathfrak{sl}_3(\mathbb{C}))$ and $\mathbb{C}$ is the class of all finite-dimensional irreducible representations of $A$ with sufficiently high weights. In case (1) the answer is: tuples that are linearly dependent over $\mathbb{C}$ while in cases (2) and (3) the answer is: tuples that are linearly dependent over the center of $A$. Similar results have been proved before for free algebras and Weyl algebras.

math.RT↗

arXiv2017

The singular bivariate quartic tracial moment problem

The (classical) truncated moment problem, extensively studied by Curto and Fialkow, asks to characterize when a finite sequence of real numbers indexes by words in commuting variables can be represented with moments of a positive Borel measure $μ$ on $\mathbb R^n$. In \cite{BK12} Burgdorf and Klep introduced its tracial analog, the truncated tracial moment problem, which replaces commuting variables with non-commuting ones and moments of $μ$ with tracial moments of matrices. In the bivariate quartic case, where indices run over words in two variables of degree at most four, every sequence with a positive definite $7\times 7$ moment matrix $\mathcal M_2$ can be represented with tracial moments \cite{BK10,BK12}. In this article the case of singular $\mathcal M_2$ is studied. For $\mathcal M_2$ of rank at most 5 the problem is solved completely; namely, concrete measures are obtained whenever they exist and the uniqueness question of the minimal measures is answered. For $\mathcal M_2$ of rank 6 the problem splits into four cases, in two of which it is equivalent to the feasibility problem of certain linear matrix inequalities. Finally, the question of a flat extension of the moment matrix $\mathcal M_2$ is addressed. While this is the most powerful tool for solving the classical case, it is shown here by examples that, while sufficient, flat extensions are mostly not a necessary condition for the existence of a measure in the tracial case.

math.FA↗

arXiv2017

There are many more positive maps than completely positive maps

A linear map between matrix spaces is positive if it maps positive semidefinite matrices to positive semidefinite ones, and is called completely positive if all its ampliations are positive. In this article quantitative bounds on the fraction of positive maps that are completely positive are proved. A main tool are real algebraic geometry techniques developed by Blekherman to study the gap between positive polynomials and sums of squares. Finally, an algorithm to produce positive maps which are not completely positive is given.

math.FA↗

arXiv2016

Operator Positivstellensätze for noncommutative polynomials positive on matrix convex sets

This article studies algebraic certificates of positivity for noncommutative (nc) operator-valued polynomials on matrix convex sets, such as the solution set $D_L$, called a free Hilbert spectrahedron, of the linear operator inequality (LOI) $L(X)=A_0\otimes I+\sum_{j=1}^g A_{j}\otimes X_j\succeq 0,$ where $A_j$ are self-adjoint linear operators on a separable Hilbert space, $X_j$ matrices and $I$ is an identity matrix. If $A_j$ are matrices, then $L(X)\succeq 0$ is called a linear matrix inequality (LMI) and $D_L$ a free spectrahedron. For monic LMIs, i.e., $A_0=I$, and nc matrix-valued polynomials the certificates of positivity were established by Helton, Klep and McCollough in a series of articles with the use of the theory of complete positivity from operator algebras and classical separation arguments from real algebraic geometry. Since the full strength of the theory of complete positivity is not restricted to finite dimensions, but works well also in the infinite-dimensional setting, we use it to tackle our problems. First we extend the characterization of the inclusion $D_{L_1}\subseteq D_{L_2}$ from monic LMIs to monic LOIs. As a corollary one obtains the description of a polar dual of a free Hilbert spectrahedron $D_L$ and its projection, called a free Hilbert spectrahedrop. Using this characterization in a separation argument, we obtain a certificate for multivariate matrix-valued nc polynomials $F$ positive semidefinite on a free Hilbert spectrahedron defined by a monic LOI. Replacing the separation argument by an operator Fejér-Riesz theorem enables us to extend this certificate, in the univariate case, to operator-valued polynomials $F$. Finally, focusing on the algebraic description of the equality $D_{L_1}=D_{L_2}$, we remove the assumption of boundedness from the description in the LMIs case by an extended analysis. However, the description does not extend to LOIs.

math.OA↗

arXiv2015

Matrix Fejér-Riesz theorem with gaps

The matrix Fejér-Riesz theorem characterizes positive semidefinite matrix polynomials on the real line $\mathbb{R}$. We extend a characterization to arbitrary closed semialgebraic sets $K\subseteq \mathbb{R}$ by the use of matrix preorderings from real algebraic geometry. In the compact case a denominator-free characterization exists, while in the non-compact case there are counterexamples. However, there is a weaker characterization with denominators in the non-compact case. At the end we extend the results to algebraic curves.

math.AG↗

arXiv2015

Integral representations of -representations of -algebras

Regular normalized $B(W_1,W_2)$-valued non-negative spectral measures introduced in \cite{Zalar2014} are in one-to-one correspondence with unital $\ast$-representations $ρ:C(X,\mathbb{C})\otimes W_1 \rightarrow W_2$, where $X$ stands for a compact Hausdorff space and $W_1, W_2$ stand for von Neumann algebras. In this paper we generalize this result in two directions. The first is to $\ast$-representations of the form $ρ:\mathcal{B}\otimes W_1\rightarrow W_2$, where $\mathcal{B}$ stands for a commutative $\ast$-algebra $\mathcal{B}$, and the second is to special (not necessarily bounded) $\ast$-representations of the form $ρ:\mathcal{B}\otimes W_1\rightarrow \mathcal{L}^+(\mathcal{D})$, where $\mathcal{L}^+(\mathcal{D})$ stands for a $\ast$-algebra of special linear operators on a dense subspace $\mathcal{D}$ of a Hilbert space $\mathcal{K}$.

math.OA↗

arXiv2014

Non-negative spectral measures and representations of C*-algebras

Regular normalized W-valued spectral measures on a compact Hausdorff space X are in one-to-one correspondence with unital *-representations ρ:C(X)\to W, where W stands for a von Neumann algebra. In this paper we show that for every compact Hausdorff space X and every von Neumann algebras W_1,W_2 there is a one-to-one correspondence between unital *-representations ρ:C(X,W_1)\to W_2 and special B(W_1,W_2)-valued measures on X that we call non-negative spectral measures. Such measures are special cases of non-negative measures that we introduced in our previous paper in connection with moment problems for operator polynomials.

math.FA↗

arXiv2012

A note on a matrix version of the Farkas lemma

A linear polyomial non-negative on the non-negativity domain of finitely many linear polynomials can be expressed as their non-negative linear combination. Recently, under several additional assumptions, Helton, Klep, and McCullough extended this result to matrix polynomials. The aim of this paper is to study which of these additional assumptions are really necessary.

math.OA↗