Showing 1–14 of 14 results for author: Pagacz, P

On the dimension of orbits of matrix pencils under strict equivalence

Authors: Fernando De Terán, Froilán M. Dopico, Patryk Pagacz

Abstract: We prove that, given two matrix pencils $L$ and $M$, if $M$ belongs to the closure of the orbit of $L$ under strict equivalence, then the dimension of the orbit of $M$ is smaller than or equal to the dimension of the orbit of $L$, and the equality is only attained when $M$ belongs to the orbit of $L$. Our proof uses only the majorization involving the eigenstructures of $L$ and $M$ which character… ▽ More We prove that, given two matrix pencils $L$ and $M$, if $M$ belongs to the closure of the orbit of $L$ under strict equivalence, then the dimension of the orbit of $M$ is smaller than or equal to the dimension of the orbit of $L$, and the equality is only attained when $M$ belongs to the orbit of $L$. Our proof uses only the majorization involving the eigenstructures of $L$ and $M$ which characterizes the inclusion relationship between orbit closures, together with the formula for the codimension of the orbit of a pencil in terms of its eigenstruture. △ Less

Submitted 4 May, 2025; originally announced May 2025.

MSC Class: 15A18; 15A21; 15A22; 15A54; 65F15

The Taylor spectrum of pairs of isometries

Authors: Zbigniew Burdak, Patryk Pagacz

Abstract: In the paper we fully describe Taylor spectrum of pairs of isometries given by diagrams. In most cases both isometries in such pairs have non-trivial shift part and its Taylor spectrum is a proper subset (of Lebesgue measure in $(0,π^2)$) of the closed bidisc. In the paper we fully describe Taylor spectrum of pairs of isometries given by diagrams. In most cases both isometries in such pairs have non-trivial shift part and its Taylor spectrum is a proper subset (of Lebesgue measure in $(0,π^2)$) of the closed bidisc. △ Less

Submitted 31 October, 2024; originally announced October 2024.

Comments: 24 pages

MSC Class: 47B02; 47B40; 47A13

An operator theory approach to the evanescent part of a two-parametric weak-stationary stochastic process

Authors: Zbigniew Burdak, Marek Kosiek, Patryk Pagacz, Marek Słociński

Abstract: A new approach to the evanescent part of a two-dimensional weak-stationary stochastic process with the past given by a half-plane is proceed. The classical result due to Helson and Lowdenslager divides a two-parametric weak-stationary stochastic process into three parts. In this paper we describe the most untouchable one - the evanescent part. Moreover, we point out how this part depends on the sh… ▽ More A new approach to the evanescent part of a two-dimensional weak-stationary stochastic process with the past given by a half-plane is proceed. The classical result due to Helson and Lowdenslager divides a two-parametric weak-stationary stochastic process into three parts. In this paper we describe the most untouchable one - the evanescent part. Moreover, we point out how this part depends on the shape of the past. △ Less

Submitted 25 July, 2024; originally announced July 2024.

Comments: 20 pages

MSC Class: 47B20; 47A13; 60G10; 60G25

On bundle closures of matrix pencils and matrix polynomials

Authors: Fernando De Terán, Froilán M. Dopico, Vadym Koval, Patryk Pagacz

Abstract: Bundles of matrix polynomials are sets of matrix polynomials with the same size and grade and the same eigenstructure up to the specific values of the eigenvalues. It is known that the closure of the bundle of a pencil $L$ (namely, a matrix polynomial of grade $1$), denoted by $\mathcal{B}(L)$, is the union of $\mathcal{B}(L)$ itself with a finite number of other bundles. The first main contributi… ▽ More Bundles of matrix polynomials are sets of matrix polynomials with the same size and grade and the same eigenstructure up to the specific values of the eigenvalues. It is known that the closure of the bundle of a pencil $L$ (namely, a matrix polynomial of grade $1$), denoted by $\mathcal{B}(L)$, is the union of $\mathcal{B}(L)$ itself with a finite number of other bundles. The first main contribution of this paper is to prove that the dimension of each of these bundles is strictly smaller than the dimension of $\mathcal{B}(L)$. The second main contribution is to prove that also the closure of the bundle of a matrix polynomial of grade larger than 1 is the union of the bundle itself with a finite number of other bundles of smaller dimension. To get these results we obtain a formula for the (co)dimension of the bundle of a matrix pencil in terms of the Weyr characteristics of the partial multiplicities of the eigenvalues and of the (left and right) minimal indices, and we provide a characterization for the inclusion relationship between the closures of two bundles of matrix polynomials of the same size and grade. △ Less

Submitted 26 February, 2024; originally announced February 2024.

Comments: 22 pages

MSC Class: 15A18; 15A21; 15A22; 15A54; 65F15

On singular pencils with commuting coefficients

Authors: Vadym Koval, Patryk Pagacz

Abstract: We investigate the relation between the spectrum of matrix (or operator) polynomials and the Taylor spectrum of its coefficients. We prove that the polynomial of commuting matrices is singular, i.e. its spectrum is the whole complex plane, if and only if (0, 0, ... , 0) belongs to the Taylor spectrum of its coefficients. On the other hand we prove that this equivalence is not longer true if we con… ▽ More We investigate the relation between the spectrum of matrix (or operator) polynomials and the Taylor spectrum of its coefficients. We prove that the polynomial of commuting matrices is singular, i.e. its spectrum is the whole complex plane, if and only if (0, 0, ... , 0) belongs to the Taylor spectrum of its coefficients. On the other hand we prove that this equivalence is not longer true if we consider the operators on infinite dimensional Hilbert space as coefficients of polynomial. As a consequence we could propose a new description of (Taylor) spectrum of k-tuple of matrices and we could disprove the conjecture previously proposed in the literature. Additionally, we pointed out the Kronecker forms of the pencils with commuting coefficients. △ Less

Submitted 16 March, 2024; v1 submitted 1 February, 2024; originally announced February 2024.

Comments: 15 pages

MSC Class: 15A22; 15A27; 47A13; 15A60

Matrix pencils with the numerical range equal to the whole complex plane

Authors: Vadym Koval, Patryk Pagacz

Abstract: The main purpose of this article is to show that the numerical range of a linear pencil $λA + B$ is equal to $\mathbb{C}$ if and only if $0$ belongs to the convex hull of the joint numerical range of $A$ and $B$. We also prove that if the numerical range of a linear pencil $λA + B$ is equal to $\mathbb{C}$ and $A + A^*, B + B^* \geq 0$, then $A$ and $B$ have a common isotropic vector. Moreover, we… ▽ More The main purpose of this article is to show that the numerical range of a linear pencil $λA + B$ is equal to $\mathbb{C}$ if and only if $0$ belongs to the convex hull of the joint numerical range of $A$ and $B$. We also prove that if the numerical range of a linear pencil $λA + B$ is equal to $\mathbb{C}$ and $A + A^*, B + B^* \geq 0$, then $A$ and $B$ have a common isotropic vector. Moreover, we improve the classical result which describes Hermitian linear pencils. △ Less

Submitted 20 May, 2022; v1 submitted 10 May, 2022; originally announced May 2022.

Comments: 9

MSC Class: 15A22; 15A60

Target Layer Regularization for Continual Learning Using Cramer-Wold Generator

Authors: Marcin Mazur, Łukasz Pustelnik, Szymon Knop, Patryk Pagacz, Przemysław Spurek

Abstract: We propose an effective regularization strategy (CW-TaLaR) for solving continual learning problems. It uses a penalizing term expressed by the Cramer-Wold distance between two probability distributions defined on a target layer of an underlying neural network that is shared by all tasks, and the simple architecture of the Cramer-Wold generator for modeling output data representation. Our strategy… ▽ More We propose an effective regularization strategy (CW-TaLaR) for solving continual learning problems. It uses a penalizing term expressed by the Cramer-Wold distance between two probability distributions defined on a target layer of an underlying neural network that is shared by all tasks, and the simple architecture of the Cramer-Wold generator for modeling output data representation. Our strategy preserves target layer distribution while learning a new task but does not require remembering previous tasks' datasets. We perform experiments involving several common supervised frameworks, which prove the competitiveness of the CW-TaLaR method in comparison to a few existing state-of-the-art continual learning models. △ Less

Submitted 15 November, 2021; originally announced November 2021.

Comments: The paper is under consideration at Computer Vision and Image Understanding

The decomposition theorems in Baer $*$-rings

Authors: Zbigniew Burdak, Marek Kosiek, Patryk Pagacz, Marek Słociński

Abstract: We show a general decomposition theorem in Baer *-rings. As a consequence the vast majority of decompositions known in the algebra of bounded Hilbert space operators are generalized to Baer *-rings. There are also results which are new in the algebra of bounded Hilbert space operators. The model of summands in Wold-Słociński decomposition in Baer *-rings is given. We show a general decomposition theorem in Baer *-rings. As a consequence the vast majority of decompositions known in the algebra of bounded Hilbert space operators are generalized to Baer *-rings. There are also results which are new in the algebra of bounded Hilbert space operators. The model of summands in Wold-Słociński decomposition in Baer *-rings is given. △ Less

Submitted 5 September, 2019; originally announced September 2019.

MSC Class: 16W10

Between the von Neumann inequality and the Crouzeix conjecture

Authors: Patryk Pagacz, Paweł Pietrzycki, Michał Wojtylak

Abstract: A new concept of a deformed numerical range $W^ρ(T)$ is introduced. Here $T$ is a bounded linear operator or a matrix and $ ρ\in[1,+\infty)$ is a parameter. Each $W^ρ(T)$ is a closed convex set that contains the spectrum of $T$. Furthermore, $W^ρ(T)$ is decreasing with respect to $ ρ$ and $W^2(T)$ coincides with the numerical range. It is also shown that $W^ρ(T)$ is contained in the closed unit di… ▽ More A new concept of a deformed numerical range $W^ρ(T)$ is introduced. Here $T$ is a bounded linear operator or a matrix and $ ρ\in[1,+\infty)$ is a parameter. Each $W^ρ(T)$ is a closed convex set that contains the spectrum of $T$. Furthermore, $W^ρ(T)$ is decreasing with respect to $ ρ$ and $W^2(T)$ coincides with the numerical range. It is also shown that $W^ρ(T)$ is contained in the closed unit disc if and only if $T$ has a $ρ$ unitary dilation in the sense of Nágy-Foia\c s. The spectral constants of $W^ρ(T)$ are investigated, it is shown that it is monotone and continuous with respect to the parameter $ ρ$. △ Less

Submitted 21 May, 2020; v1 submitted 14 July, 2019; originally announced July 2019.

Journal ref: Linear Algebra and Its Applications vol. 605 (2020), 130-157

Invariant subspaces of $\mathcal{H}^2(\mathbb{T}^2)$ and $L^2(\mathbb{T}^2)$ preserving compatibility

Authors: Zbigniew Burdak, Marek Kosiek, Patryk Pagacz, Marek Słociński

Abstract: Operators of multiplication by independent variables on the space of square summable functions over the torus and its Hardy subspace are considered. Invariant subspaces where the operators are compatible are described. Operators of multiplication by independent variables on the space of square summable functions over the torus and its Hardy subspace are considered. Invariant subspaces where the operators are compatible are described. △ Less

Submitted 14 March, 2017; originally announced March 2017.

Comments: 17 pages, 3 figures

MSC Class: 47A15 (Primary); 47B37 (Secondary)

Journal ref: J. Math. Anal. Appl., 455(2017), 1706-1719

Random Perturbations of Matrix Polynomials

Authors: Patryk Pagacz, Michał Wojtylak

Abstract: A sum of a large-dimensional random matrix polynomial and a fixed low-rank matrix polynomial is considered. The main assumption is that the resolvent of the random polynomial converges to some deterministic limit. A formula for the limit of the resolvent of the sum is derived and the eigenvalues are localised. Three instances are considered: a low-rank matrix perturbed by the Wigner matrix, a prod… ▽ More A sum of a large-dimensional random matrix polynomial and a fixed low-rank matrix polynomial is considered. The main assumption is that the resolvent of the random polynomial converges to some deterministic limit. A formula for the limit of the resolvent of the sum is derived and the eigenvalues are localised. Three instances are considered: a low-rank matrix perturbed by the Wigner matrix, a product $HX$ of a fixed diagonal matrix $H$ and the Wigner matrix $X$ and a special matrix polynomial. The results are illustrated with various examples and numerical simulations. △ Less

Submitted 14 October, 2020; v1 submitted 6 March, 2017; originally announced March 2017.

Comments: 32 pages, 6 figures

MSC Class: Primary 15A18 Secondary 15B52; 47A56

Journal ref: J Theor Probab 35, 52-88 (2022)

The Berberian's transform and an asymmetric Putnam-Fuglede theorem

Authors: Ahmed Bachir, Patryk Pagacz

Abstract: We present how to apply a Berberian's technique to asymmetric Putnam-Fuglede theorems. In particular, we proved that if $A, B \in B(H)$ belong to the union of classes of $*$-paranormal operators, p-hyponormal operators, dominant operators and operators of class Y and $AX = XB^*$ for some $X \in B(H)$, then $A^*X = XB$. Moreover, we gave a new counterexample for an asymmetric Putnam-Fuglede theorem… ▽ More We present how to apply a Berberian's technique to asymmetric Putnam-Fuglede theorems. In particular, we proved that if $A, B \in B(H)$ belong to the union of classes of $*$-paranormal operators, p-hyponormal operators, dominant operators and operators of class Y and $AX = XB^*$ for some $X \in B(H)$, then $A^*X = XB$. Moreover, we gave a new counterexample for an asymmetric Putnam-Fuglede theorem for paranormal operators △ Less

Submitted 23 August, 2021; v1 submitted 19 May, 2014; originally announced May 2014.

Comments: 11 pages

MSC Class: 47B20; 47A30; 47B47

Journal ref: Complex Anal. Oper. Theory 16, 59 (2022)

On the spectral properties of a class of $H$-selfadjoint random matrices and the underlying combinatorics

Authors: Patryk Pagacz, Michal Wojtylak

Abstract: An expansion of the Weyl function of a $H$-selfadjoint random matrix with one negative square is provided. It is shown that the coefficients converge to a certain generalization of Catlan numbers. Properties of this generalization are studied, in particular, a combinatorial interpretation is given. An expansion of the Weyl function of a $H$-selfadjoint random matrix with one negative square is provided. It is shown that the coefficients converge to a certain generalization of Catlan numbers. Properties of this generalization are studied, in particular, a combinatorial interpretation is given. △ Less

Submitted 8 October, 2013; originally announced October 2013.

MSC Class: Primary 15B52; Secondary 15B57; 05A19

On the power-bounded operators of classes $C_{0 \cdot}$ and $C_{1 \cdot}$

Authors: Patryk Pagacz

Abstract: By a bounded backward sequence of the operator $T$ we mean a bounded sequence $\{x_n\}$ satisfying $Tx_{n+1}=x_n$. In \cite{Pa} we have characterized contractions with strongly stable nonunitary part in terms of bounded backward sequences. The main purpose of this work is to extend that result to power-bounded operators. Aditionally, we show that a power-bounded operator is strongly stable (… ▽ More By a bounded backward sequence of the operator $T$ we mean a bounded sequence $\{x_n\}$ satisfying $Tx_{n+1}=x_n$. In \cite{Pa} we have characterized contractions with strongly stable nonunitary part in terms of bounded backward sequences. The main purpose of this work is to extend that result to power-bounded operators. Aditionally, we show that a power-bounded operator is strongly stable ($C_{0 \cdot} $) if and only if its adjoint does not have any nonzero bounded backward sequence. Similarly, a power-bounded operator is non-vanishing ($C_{1 \cdot} $) if and only if its adjoint has a lot of bounded backward sequences. △ Less

Submitted 3 June, 2012; originally announced June 2012.

MSC Class: 47A05; 47A45; 47B37; 47G10