arXiv subjects

Palle Jorgensen

Publications and source records attributed to Palle Jorgensen.

At least 55 records · Page 3Linked to original sources

arXiv2015

A new realization of rational functions, with applications to linear combination interpolation

We introduce the following linear combination interpolation problem (LCI): Given $N$ distinct numbers $w_1,\ldots w_N$ and $N+1$ complex numbers $a_1,\ldots, a_N$ and $c$, find all functions $f(z)$ analytic in a simply connected set (depending on $f$) containing the points $w_1,\ldots,w_N$ such that \[ \sum_{u=1}^Na_uf(w_u)=c. \] To this end we prove a representation theorem for such functions $f$ in terms of an associated polynomial $p(z)$. We first introduce the following two operations, $(i)$ substitution of $p$, and $(ii)$ multiplication by monomials $z^j, 0\le j < N$. Then let $M$ be the module generated by these two operations, acting on functions analytic near $0$. We prove that every function $f$, analytic in a neighborhood of the roots of $p$, is in $M$. In fact, this representation of $f$ is unique. To solve the above interpolation problem, we employ an adapted systems theoretic realization, as well as an associated representation of the Cuntz relations (from multi-variable operator theory.) We study these operations in reproducing kernel Hilbert space): We give necessary and sufficient condition for existence of realizations of these representation of the Cuntz relations by operators in certain reproducing kernel Hilbert spaces, and offer infinite product factorizations of the corresponding kernels.

math.FA↗

arXiv2015

Infinite networks and variation of conductance functions in discrete Laplacians

For a given infinite connected graph $G=(V,E)$ and an arbitrary but fixed conductance function $c$, we study an associated graph Laplacian $Δ_{c}$; it is a generalized difference operator where the differences are measured across the edges $E$ in $G$; and the conductance function $c$ represents the corresponding coefficients. The graph Laplacian (a key tool in the study of infinite networks) acts in an energy Hilbert space $\mathscr{H}_{E}$ computed from $c$. Using a certain Parseval frame, we study the spectral theoretic properties of graph Laplacians. In fact, for fixed $c$, there are two versions of the graph Laplacian, one defined naturally in the $l^{2}$ space of $V$, and the other in $\mathscr{H}_{E}$. The first is automatically selfadjoint, but the second involves a Krein extension. We prove that, as sets, the two spectra are the same, aside from the point 0. The point zero may be in the spectrum of the second, but not the first. We further study the fine structure of the respective spectra as the conductance function varies; showing now how the spectrum changes subject to variations in the function $c$.

math.FA↗

arXiv2015

Infinite weighted graphs with bounded resistance metric

We consider infinite weighted graphs $G$, i.e., sets of vertices $V$, and edges $E$ assumed countable infinite. An assignment of weights is a positive symmetric function $c$ on $E$ (the edge-set), conductance. From this, one naturally defines a reversible Markov process, and a corresponding Laplace operator acting on functions on $V$, voltage distributions. The harmonic functions are of special importance. We establish explicit boundary representations for the harmonic functions on $G$ of finite energy. We compute a resistance metric $d$ from a given conductance function. (The resistance distance $d(x,y)$ between two vertices $x$ and $y$ is the voltage drop from $x$ to $y$, which is induced by the given assignment of resistors when 1 amp is inserted at the vertex $x$, and then extracted again at $y$.) We study the class of models where this resistance metric is bounded. We show that then the finite-energy functions form an algebra of $\frac{1}{2}$-Lipschitz-continuous and bounded functions on $V$, relative to the metric $d$. We further show that, in this case, the metric completion $M$ of $(V,d)$ is automatically compact, and that the vertex-set $V$ is open in $M$. We obtain a Poisson boundary-representation for the harmonic functions of finite energy, and an interpolation formula for every function on $V$ of finite energy. We further compare $M$ to other compactifications; e.g., to certain path-space models.

math.FA↗

arXiv2015

Realizations of infinite products, Ruelle operators and wavelet filters

Using the notions and tools from realization in the sense of systems theory, we establish an explicit and new realization formula for families of infinite products of rational matrix-functions of a single complex variable. Our realizations of these resulting infinite products have the following four features: 1) Our infinite product realizations are functions defined in an infinite-dimensional complex domain. 2) Starting with a realization of a single rational matrix-function $M$, we show that a resulting infinite product realization obtained from $M$ takes the form of an (infinite-dimensional) Toeplitz operator with a symbol that is a reflection of the initial realization for $M$. 3) Starting with a subclass of rational matrix functions, including scalar-valued corresponding to low-pass wavelet filters, we obtain the corresponding infinite products that realize the Fourier transforms of generators of $\mathbf L_2(\mathbb R)$ wavelets. 4) We use both the realizations for $M$ and the corresponding infinite product to produce a matrix representation of the Ruelle-transfer operators used in wavelet theory. By matrix representation we refer to the slanted (and sparse) matrix which realizes the Ruelle-transfer operator under consideration.

math.CV↗

arXiv2015

Generalized Gramians: Creating frame vectors in maximal subspaces

A frame is a system of vectors $S$ in Hilbert space $\mathscr{H}$ with properties which allow one to write algorithms for the two operations, analysis and synthesis, relative to $S$, for all vectors in $\mathscr{H}$; expressed in norm-convergent series. Traditionally, frame properties are expressed in terms of an $S$-Gramian, $G_{S}$ (an infinite matrix with entries equal to the inner product of pairs of vectors in $S$); but still with strong restrictions on the given system of vectors in $S$, in order to guarantee frame-bounds. In this paper we remove these restrictions on $G_{S}$, and we obtain instead direct-integral analysis/synthesis formulas. We show that, in spectral subspaces of every finite interval $J$ in the positive half-line, there are associated standard frames, with frame-bounds equal the endpoints of $J$. Applications are given to reproducing kernel Hilbert spaces, and to random fields.

math.FA↗

arXiv2015

Discrete reproducing kernel Hilbert spaces: Sampling and distribution of Dirac-masses

We study reproducing kernels, and associated reproducing kernel Hilbert spaces (RKHSs) $\mathscr{H}$ over infinite, discrete and countable sets $V$. In this setting we analyze in detail the distributions of the corresponding Dirac point-masses of $V$. Illustrations include certain models from neural networks: An Extreme Learning Machine (ELM) is a neural network-configuration in which a hidden layer of weights are randomly sampled, and where the object is then to compute resulting output. For RKHSs $\mathscr{H}$ of functions defined on a prescribed countable infinite discrete set $V$, we characterize those which contain the Dirac masses $δ_{x}$ for all points $x$ in $V$. Further examples and applications where this question plays an important role are: (i) discrete Brownian motion-Hilbert spaces, i.e., discrete versions of the Cameron-Martin Hilbert space; (ii) energy-Hilbert spaces corresponding to graph-Laplacians where the set $V$ of vertices is then equipped with a resistance metric; and finally (iii) the study of Gaussian free fields.

math.FA↗

arXiv2015

Characterizations of rectangular (para)-unitary rational Functions

We here present three characterizations of not necessarily causal, rational functions which are (co)-isometric on the unit circle: (i) Through the realization matrix of Schur stable systems. (ii) The Blaschke-Potapov product, which is then employed to introduce an easy-to-use description of all these functions with dimensions and McMillan degree as parameters. (iii) Through the (not necessarily reducible) Matrix Fraction Description (MFD). In cases (ii) and (iii) the poles of the rational functions involved may be anywhere in the complex plane, but the unit circle (including both zero and infinity). A special attention is devoted to exploring the gap between the square and rectangular cases.

math.CV↗

arXiv2014

Reproducing kernel Hilbert spaces generated by the binomial coefficients

We study a reproducing kernel Hilbert space of functions defined on the positive integers and associated to the binomial coefficients. We introduce two transforms, which allow us to develop a related harmonic analysis in this Hilbert space. Finally, we mention connections with the theory of discrete analytic functions and with the quantum case.

math.CV↗

arXiv2014

Filters and Matrix Factorization

We give a number of explicit matrix-algorithms for analysis/synthesis in multi-phase filtering; i.e., the operation on discrete-time signals which allow a separation into frequency-band components, one for each of the ranges of bands, say $N$, starting with low-pass, and then corresponding filtering in the other band-ranges. If there are $N$ bands, the individual filters will be combined into a single matrix action; so a representation of the combined operation on all $N$ bands by an $N \times N$ matrix, where the corresponding matrix-entries are periodic functions; or their extensions to functions of a complex variable. Hence our setting entails a fixed $N \times N$ matrix over a prescribed algebra of functions of a complex variable. In the case of polynomial filters, the factorizations will always be finite. A novelty here is that we allow for a wide family of non-polynomial filter-banks. Working modulo $N$ in the time domain, our approach also allows for a natural matrix-representation of both down-sampling and up-sampling. The implementation encompasses the combined operation on input, filtering, down-sampling, transmission, up-sampling, an action by dual filters, and synthesis, merges into a single matrix operation. Hence our matrix-factorizations break down the global filtering-process into elementary steps. To accomplish this, we offer a number of adapted matrix factorization-algorithms, such that each factor in our product representation implements in a succession of steps the filtering across pairs of frequency-bands; and so it is of practical significance in implementing signal processing, including filtering of digitized images. Our matrix-factorizations are especially useful in the case of the processing a fixed, but large, number of bands.

math.FA↗

arXiv2014

Characterizations of families of rectangular finite impulse response, para-unitary systems

We here study Finite Impulse Response (FIR) rectangular, not necessarily causal, systems which are (para)-unitary on the unit circle (=the class $\mathcal{U}$). First, we offer three characterizations of these systems. Then, introduce a %easy-to-use description of all FIRs in $\mathcal{U}$, as copies of a real polytope, parametrized by the dimensions and the McMillan degree of the FIRs. Finally, we present six simple ways (along with their combinations) to construct, from any FIR, a large family of FIRs, of various dimensions and McMillan degrees, so that whenever the original system is in $\mathcal{U}$, so is the whole family. A key role is played by Hankel matrices.

math.CV↗

arXiv2014

Unbounded operators, Lie algebras, and local representations

We prove a number of results on integrability and extendability of Lie algebras of unbounded skew-symmetric operators with common dense domain in Hilbert space. By integrability for a Lie algebra $\mathfrak{g}$, we mean that there is an associated unitary representation $\mathcal{U}$ of the corresponding simply connected Lie group such that $\mathfrak{g}$ is the differential of $\mathcal{U}$. Our results extend earlier integrability results in the literature; and are new even in the case of a single operator. Our applications include a new invariant for certain Riemann surfaces.

math.FA↗

arXiv2014

Frames and Factorization of Graph Laplacians

Using functions from electrical networks (graphs with resistors assigned to edges), we prove existence (with explicit formulas) of a canonical Parseval frame in the energy Hilbert space $\mathscr{H}_{E}$ of a prescribed infinite (or finite) network. Outside degenerate cases, our Parseval frame is not an orthonormal basis. We apply our frame to prove a number of explicit results: With our Parseval frame and related closable operators in $\mathscr{H}_{E}$ we characterize the Friedrichs extension of the $\mathscr{H}_{E}$-graph Laplacian. We consider infinite connected network-graphs $G=\left(V,E\right)$, $V$ for vertices, and \emph{E} for edges. To every conductance function $c$ on the edges $E$ of $G$, there is an associated pair $\left(\mathscr{H}_{E},Δ\right)$ where $\mathscr{H}_{E}$ in an energy Hilbert space, and $Δ\left(=Δ_{c}\right)$ is the $c$-Graph Laplacian; both depending on the choice of conductance function $c$. When a conductance function is given, there is a current-induced orientation on the set of edges and an associated natural Parseval frame in $\mathscr{H}_{E}$ consisting of dipoles. Now $Δ$ is a well-defined semibounded Hermitian operator in both of the Hilbert $l^{2}\left(V\right)$ and $\mathscr{H}_{E}$. It is known to automatically be essentially selfadjoint as an $l^{2}\left(V\right)$-operator, but generally not as an $\mathscr{H}_{E}$ operator. Hence as an $\mathscr{H}_{E}$ operator it has a Friedrichs extension. In this paper we offer two results for the Friedrichs extension: a characterization and a factorization. The latter is via $l^{2}\left(V\right)$.

math.FA↗

arXiv2014

von Neumann indices and classes of positive definite functions

With view to applications, we establish a correspondence between two problems: (i) the problem of finding continuous positive definite extensions of functions $F$ which are defined on open bounded domains $Ω$ in $\mathbb{R}$, on the one hand; and (ii) spectral theory for elliptic differential operators acting on $Ω$, (constant coefficients.) A novelty in our approach is the use of a reproducing kernel Hilbert space $\mathscr{H}_{F}$ computed directly from $\left(Ω,F\right)$, as well as algorithms for computing relevant orthonormal bases in $\mathscr{H}_{F}$.

math.FA↗

arXiv2014

Harmonic analysis of a class of reproducing kernel Hilbert spaces arising from groups

We study two extension problems, and their interconnections: (i) extension of positive definite (p.d.) continuous functions defined on subsets in locally compact groups $G$; and (ii) (in case of Lie groups $G$) representations of the associated Lie algebras $La\left(G\right)$, i.e., representations of $La\left(G\right)$ by unbounded skew-Hermitian operators acting in a reproducing kernel Hilbert space $\mathscr{H}_{F}$ (RKHS). Our analysis is non-trivial even if $G=\mathbb{R}^{n}$, and even if $n=1$. If $G=\mathbb{R}^{n}$, (ii), we are concerned with finding systems of strongly commuting selfadjoint operators $\left\{ T_{i}\right\} $ extending a system of commuting Hermitian operators with common dense domain in $\mathscr{H}_{F}$. Specifically, we consider partially defined positive definite (p.d.) continuous functions $F$ on a fixed group. From $F$ we then build a reproducing kernel Hilbert space $\mathscr{H}_{F}$, and the operator extension problem is concerned with operators acting in $\mathscr{H}_{F}$, and with unitary representations of $G$ acting on $\mathscr{H}_{F}$. Our emphasis is on the interplay between the two problems, and on the harmonic analysis of our RKHSs $\mathscr{H}_{F}$.

math.FA↗

arXiv2013

Extension of positive definite functions

Let $Ω\subset\mathbb{R}^n$ be an open, connected subset of $\mathbb{R}^n$, and let $F\colonΩ-Ω\to\mathbb{C}$, where $Ω-Ω=\{x-y\colon x,y\inΩ\}$, be a continuous positive definite function. We give necessary and sufficient conditions for $F$ to have an extension to a continuous positive definite function defined on the entire Euclidean space $\mathbb{R}^n$. The conditions are formulated in terms of strong commutativity of a system of certain unbounded selfadjoint operators defined on a Hilbert space associated to $F$. When a positive definite function $F$ is extendable, we show that it is characterized by existence of associated unitary representations of $\mathbb{R}^n$. Different positive definite extensions correspond to different unitary representations. We prove that each such unitary representation has simple spectrum. We give necessary and sufficient conditions for a continuous positive definite function to have exactly one extension. Our proof regarding extensions of positive definite functions carries over mutatis mutandis to the case of conditionally negative definite functions, which has applications to Gaussian stochastic processes, whose increments in mean-square are stationary (e.g., fractional Brownian motion).

math.SP↗

arXiv2013

On free stochastic processes and their derivatives

We study a family of free stochastic processes whose covariance kernels $K$ may be derived as a transform of a tempered measure $σ$. These processes arise, for example, in consideration non-commutative analysis involving free probability. Hence our use of semi-circle distributions, as opposed to Gaussians. In this setting we find an orthonormal bases in the corresponding non-commutative $L^2$ of sample-space. We define a stochastic integral for our family of free processes.

math.OA↗

arXiv2013

Easy-to-compute parameterizations of all wavelet filters: input-output and state-space

We here use notions from the theory linear shift-invariant dynamical systems to provide an easy-to-compute characterization of all rational wavelet filters. For a given N bigger or equql to 2, the number of inputs, the construction is based on a factorization to an elementary wavelet filter along with of m elementary unitary matrices. We shall call this m the index of the filter. It turns out that the resulting wavelet filter is of McMillan degree $N((N-1)/2+m). Rational wavelet filters bounded at infinity, admit state space realization. The above input-output parameterization is exploited for a step-by-step construction (where in each the index m is increased by one) of state space model of wavelet filters.

math.CV↗

arXiv2012

On discrete analytic functions: Products, Rational Functions, and some Associated Reproducing Kernel Hilbert Spaces

We introduce a family of discrete analytic functions, called expandable discrete analytic functions, which includes discrete analytic polynomials, and define two products in this family. The first one is defined in a way similar to the Cauchy-Kovalevskaya product of hyperholomorphic functions, and allows us to define rational discrete analytic functions. To define the second product we need a new space of entire functions which is contractively included in the Fock space. We study in this space some counterparts of Schur analysis.

math.FA↗