SearcharxivSearch

arXiv subjects

Yongdo Lim

Publications and source records attributed to Yongdo Lim.

At least 19 recordsLinked to original sources

Mostow-Type Decompositions and Geometric Means for Symmetric Cones

Motivated by Mostow's decomposition theorem for positive definite matrices and by subsequent matrix factorizations involving geometric means, we establish a Mostow-type decomposition for arbitrary symmetric cones. As a consequence, we derive an analogous decomposition for the automorphism group of the cone. We also introduce a natural Hadamard metric on the symmetric cone by pulling back an $\ell^2$-product metric, together with its midpoint operation and associated Karcher mean.

math.MG

Perturbation of monic matrix polynomials

In this paper, we study the stability of matrix polynomials under structured perturbations of their coefficients. More precisely, we consider a family of matrix polynomials \[ P_u(\lambda)=A_d(u)\lambda^d+A_{d-1}(u)\lambda^{d-1}+\cdots+A_0(u), \] whose matrix coefficients depend continuously and semialgebraically on a parameter vector $u\in\mathbb{C}^p$. Assuming that the matrix polynomial is monic, we show that the spectrum, the $\varepsilon$-pseudospectrum, the numerical range, and the joint numerical range associated with $P_u(\lambda)$ define set-valued maps that are H\"older continuous with respect to the parameter $u$. Moreover, the parameter space $\mathbb{C}^p$ can be decomposed into a finite union of analytic semialgebraic submanifolds such that, on each submanifold, the eigenvalues and the Jordan pairs of $P_u(\lambda)$ depend analytically on $u$. We also note that most of the results remain valid if the monicity assumption is replaced by the local nonsingularity of the leading coefficient matrix $A_d(u)$. However, the monic setting is adopted throughout the paper in order to simplify the exposition and to avoid additional technical assumptions, which are required in particular for results concerning numerical ranges.

math.RA

Solving a Class of Nonconvex Quadratic Programs by Inertial DC Algorithms

Two inertial DC algorithms for indefinite quadratic programs under linear constraints (IQPs) are considered in this paper. Using a qualification condition related to the normal cones of unbounded pseudo-faces of the polyhedral convex constraint set, the recession cones of the corresponding faces, and the quadratic form describing the objective function, we prove that the iteration sequences in question are bounded if the given IQP has a finite optimal value. Any cluster point of such a sequence is a KKT point. The convergence of the members of a DCA sequence produced by one of the two inertial algorithms to just one connected component of the KKT point set is also obtained. To do so, we revisit the inertial algorithm for DC programming of de Oliveira and Tcheou [de Oliveira, W., Tcheou, M.P.: An inertial algorithm for DC programming, Set-Valued and Variational Analysis 2019; 27: 895--919] and give a refined version of Theorem 1 from that paper, which can be used for IQPs with unbounded constraint sets. An illustrative example is proposed.

math.OC

Local Error Bounds for Affine Variational Inequalities on Hilbert Spaces

This paper gives some results related to the research problem about infinite-dimensional affine variational inequalities raised by N.D. Yen and X. Yang [Affine variational inequalities on normed spaces, J. Optim. Theory Appl., 178 (2018), 36--55]. Namely, we obtain local error bounds for affine variational inequalities on Hilbert spaces. To do so, we revisit two fundamental properties of polyhedral mappings. Then, we prove a locally upper Lipschitzian property of the inverse of the residual mapping of the infinite-dimensional affine variational inequality under consideration. Finally, we derive the desired local error bounds from that locally upper Lipschitzian property.

math.OC

The Expanding Universe of the Geometric Mean

In this paper the authors seek to trace in an accessible fashion the rapid recent development of the theory of the matrix geometric mean in the cone of positive definite matrices up through the closely related operator geometric mean in the positive cone of a unital $C^*$-algebra. The story begins with the two-variable matrix geometric mean, moves to the breakthrough developments in the multivariable matrix setting, the main focus of the paper, and then on to the extension to the positive cone of the $C^*$-algebra of operators on a Hilbert space, even to general unital $C^*$-algebras, and finally to the consideration of barycentric maps that grow out of the geometric mean on the space of integrable probability measures on the positive cone. Besides expected tools from linear algebra and operator theory, one observes a surprisingly substantial interplay with geometrical notions in metric spaces, particularly the notion of nonpositive curvature. Added features include a glance at the probabilistic theory of random variables with values in a metric space of nonpositive curvature, and the appearance of related means such as the inductive and power means.

math.OA

Strong law of large numbers for the $L^1$-Karcher mean

Sturm's strong law of large numbers in $\mathrm{CAT}(0)$ spaces has been an influential tool to study the geometric mean or also called Karcher barycenter of positive definite matrices. It provides an easily computable stochastic approximation based on inductive means. Convergence of a deterministic version of this approximation has been proved by Holbrook, providing his "nodice" theorem for the Karcher mean of positive definite matrices. The Karcher mean has also been extended to the infinite dimensional case of positive operators on a Hilbert space by Lawson-Lim and then to probability measures with bounded support by the second author, however the $\mathrm{CAT}(0)$ property of the space is lost and one defines the mean as the unique solution of a nonlinear operator equation on a convex Banach-Finsler manifold. The formulations of Sturm's strong law of large numbers and Holbrook's "nodice" approximation are natural and both conjectured to converge, however all previous techniques of their proofs break down, due to the Banach-Finsler nature of the space. In this paper we prove both conjectures by establishing the most general $L^1$-form of Sturm's strong law of large numbers and Holbrook's "nodice" theorem in the operator norm by developing a stochastic discrete-time resolvent flow for the Karcher barycenter using its Wasserstein contraction property.

math.FA

Dirichlet forms and ultrametric Cantor sets associated to higher-rank graphs

The aim of this paper is to study the heat kernel and jump kernel of the Dirichlet form associated to ultrametric Cantor sets $\partial\BB_Λ$ that is the infinite path space of the stationary $k$-Bratteli diagram $\BB_Λ$, where $Λ$ is a finite strongly connected $k$-graph. The Dirichlet form which we are interested in is induced by an even spectral triple $(C_{\operatorname{Lip}}(\PB_Λ), π_ϕ, \mathcal{H}, D, Γ)$ and is given by \[ Q_s(f,g)=\frac{1}{2} \int_Ξ \operatorname{Tr}\big(\vert D\vert^{-s} [D,π_ϕ(f)]^{\ast} [D,π_ϕ(g)] \big) \, dν(ϕ), \] where $Ξ$ is the space of choice functions on $\partial \BB_Λ\times \partial \BB_Λ$. There are two ultrametrics, $d^{(s)}$ and $d_{w_δ}$, on $\partial \BB_Λ$ which make the infinite path space $\PB_Λ$ an ultrametric Cantor set. The former $d^{(s)}$ is associated to the eigenvalues of Laplace-Beltrami operator $Δ_s$ associated to $Q_s$, and the latter $d_{w_δ}$ is associated to a weight function $w_δ$ on $\BB_Λ$, where $δ\in (0,1)$. We show that the Perron-Frobenius measure $μ$ on $\partial \BB_Λ$ has the volume doubling property with respect to both $d^{(s)}$ and $d_{w_δ}$ and we study the asymptotic behaviors of the heat kernel associated to $Q_s$. Moreover, we show that the Dirichlet form $Q_s$ coincides with a Dirichlet form $\mathcal{Q}_{J_s, μ}$ which is associated to a jump kernel $J_s$ and the measure $μ$ on $\partial \BB_Λ$, and we investigate the asymptotic behavior and moments of displacements of the process.

math.PR

Operator means of probability measures

Let $\mathbb{P}$ be the complete metric space consisting of positive invertible operators on an infinite-dimensional Hilbert space with the Thompson metric. We introduce the notion of operator means of probability measures on $\mathbb{P}$, in parallel with Kubo and Ando's definition of two-variable operator means, and show that every operator mean is contractive for the $\infty$-Wasserstein distance. By means of a fixed point method we consider deformation of such operator means, and show that the deformation of any operator mean becomes again an operator mean in our sense. Based on this deformation procedure we prove a number of properties and inequalities for operator means of probability measures.

math.FA

Existence and uniqueness of the $L^1$-Karcher mean

We extend the domain of the Karcher mean $Λ$ of positive operators on a Hilbert space to $L^1$-Borel probability measures on the cone of positive operators equipped with the Thompson part metric. We establish existence and uniqueness of $Λ$ as the solution of the Karcher equation and develop a nonlinear ODE theory for the relative operator entropy in the spirit of Crandall-Liggett, such that the solutions of the Karcher equation are stationary solutions of the ODE, and all generated solution curves enjoy the exponential contraction estimate. This is possible despite the facts that the Thompson metric is non-Euclidean, non-differentiable, non-commutative as a metric space as well as non-separable, and the positive cone is non-locally compact as a manifold. As further applications of the ODE approach, we prove the norm convergence conjecture of the power means of positive operators to the Karcher mean, and a Trotter-Kato product formula for the nonlinear semigroups explicitly expressed by compositions of two-variable geometric means. This can be regarded as a nonlinear continuous-time version of the law of large numbers.

math.FA

Convergence of a Solution Algorithm in Indefinite Quadratic Programming

It is proved that, for an indefinite quadratic programming problem under linear constraints, any iterative sequence generated by the Proximal DC decomposition algorithm $R$-linearly converges to a Karush-Kuhn-Tucker point, provided that the problem has a solution. Another major result of this paper says that DCA sequences generated by the algorithm converge to a locally unique solution of the problem if the initial points are taken from a suitably-chosen neighborhood of it. To deal with the implicitly defined iterative sequences, a local error bound for affine variational inequalities and novel techniques are used. Numerical results together with an analysis of the influence of the decomposition parameter, as well as a comparison between the Proximal DC decomposition algorithm and the Projection DC decomposition algorithm, are given in this paper. Our results complement a recent and important paper of Le Thi, Huynh, and Pham Dinh (J. Optim. Theory Appl. 179 (2018), 103-126).

math.OC

Strong Convexity of Sandwiched Entropies and Related Optimization Problems

We present several theorems on strict and strong convexity, and higher order differential formulae for sandwiched quasi-relative entropy (a parametrised version of the classical fidelity). These are crucial for establishing global linear convergence of the gradient projection algorithm for optimisation problems for these functions. The case of the classical fidelity is of special interest for the multimarginal optimal transport problem (the $n$-coupling problem) for Gaussian measures.

math.FA

Convergence theorems for barycentric maps

We first develop a theory of conditional expectations for random variables with values in a complete metric space $M$ equipped with a contractive barycentric map $β$, and then give convergence theorems for martingales of $β$-conditional expectations. We give the Birkhoff ergodic theorem for $β$-values of ergodic empirical measures and provide a description of the ergodic limit function in terms of the $β$-conditional expectation. Moreover, we prove the continuity property of the ergodic limit function by finding a complete metric between contractive barycentric maps on the Wasserstein space of Borel probability measures on $M$. Finally, the large derivation property of $β$-values of i.i.d. empirical measures is obtained by applying the Sanov large deviation principle.

math.PR

On the Bures-Wasserstein distance between positive definite matrices

The metric $d(A,B)=\left[ \tr\, A+\tr\, B-2\tr(A^{1/2}BA^{1/2})^{1/2}\right]^{1/2}$ on the manifold of $n\times n$ positive definite matrices arises in various optimisation problems, in quantum information and in the theory of optimal transport. It is also related to Riemannian geometry. In the first part of this paper we study this metric from the perspective of matrix analysis, simplifying and unifying various proofs. Then we develop a theory of a mean of two, and a barycentre of several, positive definite matrices with respect to this metric. We explain some recent work on a fixed point iteration for computing this Wasserstein barycentre. Our emphasis is on ideas natural to matrix analysis.

math.FA

Log-majorizations for the (symplectic) eigenvalues of the Cartan barycenter

In this paper we show that the eigenvalue map and the symplectic eigenvalue map of positive definite matrices are Lipschitz for the Cartan-Hadamard Riemannian metric, and establish log-majorizations for the (symplectic) eigenvalues of the Cartan barycenter of integrable probability Borel measures. This leads a version of Jensen's inequality for geometric integrals of matrix-valued integrable random variables.

math.FA

The stochastic order of probability measures on ordered metric spaces

The general notion of a stochastic ordering is that one probability distribution is smaller than a second one if the second attaches more probability to higher values than the first. Motivated by recent work on barycentric maps on spaces of probability measures on ordered Banach spaces, we introduce and study a stochastic order on the space of probability measures $\mathcal{P}(X)$, where $X$ is a metric space equipped with a closed partial order, and derive several useful equivalent versions of the definition. We establish the antisymmetry and closedness of the stochastic order (and hence that it is a closed partial order) for the case of a partial order on a Banach space induced by a closed normal cone with interior. We also consider order-completeness of the stochastic order for a cone of a finite-dimensional Banach space and derive a version of the arithmetic-geometric-harmonic mean inequalities in the setting of the associated probability space on positive matrices.

math.FA

Geometric mean flows and the Cartan barycenter on the Wasserstein space over positive definite matrices

We introduce a class of flows on the Wasserstein space of probability measures with finite first moment on the Cartan-Hadamard Riemannian manifold of positive definite matrices, and consider the problem of differentiability of the corresponding Cartan barycentric trajectory. As a consequence we have a version of Lie-Trotter formula and a related unitarily invariant norm inequality. Furthermore, a fixed point theorem related to the Karcher equation and the Cartan barycentric trajectory is also presented as an application.

math.FA

Log-majorization and Lie-Trotter formula for the Cartan barycenter on probability measure spaces

We extend Ando-Hiai's log-majorization for the weighted geometric mean of positive definite matrices into that for the Cartan barycenter in the general setting of probability measures on the Riemannian manifold of positive definite matrices equipped with trace metric. The main key is the settlement of the monotonicity problem of the Cartan barycenteric map on the space of probability measures with finite first moment for the stochastic order induced by the cone. We also derive a version of Lie-Trotter formula and related unitarily invariant norm inequalities for the Cartan barycenter as the main application of log-majorization.

math.FA