Source author record

Jan Vybíral

Jan Vybíral appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

math.FA Information Theory math.IT math.NA math.ST Statistics Theory Machine Learning Numerical Analysis

Catalog footprint

What is connected

10works

8topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2020arXiv

A variant of Schur's product theorem and its applications

We show the following version of the Schur's product theorem. If $M=(M_{j,k})_{j,k=1}^n\in{\mathbb R}^{n\times n}$ is a positive semidefinite matrix with all entries on the diagonal equal to one, then the matrix $N=(N_{j,k})_{j,k=1}^n$ with the entries $N_{j,k}=M_{j,k}^2-\frac{1}{n}$ is positive semidefinite. As a corollary of this result, we prove the conjecture of E. Novak on intractability of numerical integration on a space of trigonometric polynomials of degree at most one in each variable. Finally, we discuss also some consequences for Bochner's theorem, covariance matrices of $χ^2$-variables, and mean absolute values of trigonometric polynomials. -- Please, have a look into page 6 of the preprint "Lower Bounds for the Error of Quadrature Formulas for Hilbert Spaces" for a discussion of the relation of Theorem 1 and Corollary 2 to Gegenbauer polynomials (pointed out by Dmitriy Bilyk (University of Minnesota)). --

preprint2016arXiv

Franke-Jawerth embeddings for Besov and Triebel-Lizorkin spaces with variable exponents

The classical Jawerth and Franke embeddings $$ F^{s_0}_{p_0,q}({\mathbb R}^n)\hookrightarrow B^{s_1}_{p_1,p_0}({\mathbb R}^n) \quad \mbox{and} \quad B^{s_0}_{p_0,p_1}({\mathbb R}^n)\hookrightarrow F^{s_1}_{p_1,q}({\mathbb R}^n) $$ are versions of Sobolev embedding between the scales of Besov and Triebel-Lizorkin function spaces for $s_0>s_1$ and $$ s_0-\frac{n}{p_0} = s_1-\frac{n}{p_1}.$$ We prove Jawerth and Franke embeddings for the scales of Besov and Triebel-Lizorkin spaces with all exponents variable $$ F^{s_0(\cdot)}_{p_0(\cdot),q(\cdot)}\hookrightarrow B^{s_1(\cdot)}_{p_1(\cdot),p_0(\cdot)} \quad \mbox{and} \quad B^{s_0(\cdot)}_{p_0(\cdot),p_1(\cdot)}\hookrightarrow F^{s_1(\cdot)}_{p_1(\cdot),q(\cdot)}, $$ respectively, if $\inf_{x\in\mathbb{R}^n}(s_0(x)-s_1(x))>0$ and $$ s_0(x) -\frac{n}{p_0(x)} = s_1(x) -\frac{n}{p_1(x)}, \quad x \in {\mathbb R}^n. $$ We work exclusively with the associated sequence spaces $b^{s(\cdot)}_{p(\cdot),q(\cdot)}$ and $f^{s(\cdot)}_{p(\cdot),q(\cdot)}$, which is justified by well known decomposition techniques. We give also a different proof of the Franke embedding in the constant exponent case which avoids duality arguments and interpolation. Our results hold also for 2-microlocal function spaces $B^{\mathbf{w}}_{p(\cdot),q(\cdot)}({\mathbb R}^n)$ and $F^{\mathbf{w}}_{p(\cdot),q(\cdot)}({\mathbb R}^n)$ which unify the smoothness scales of spaces of variable smoothness and generalized smoothness spaces.

preprint2015arXiv

Non-asymptotic Analysis of $\ell_1$-norm Support Vector Machines

Support Vector Machines (SVM) with $\ell_1$ penalty became a standard tool in analysis of highdimensional classification problems with sparsity constraints in many applications including bioinformatics and signal processing. Although SVM have been studied intensively in the literature, this paper has to our knowledge first non-asymptotic results on the performance of $\ell_1$-SVM in identification of sparse classifiers. We show that a $d$-dimensional $s$-sparse classification vector can be (with high probability) well approximated from only $O(s\log(d))$ Gaussian trials. The methods used in the proof include concentration of measure and probability in Banach spaces.

preprint2015arXiv

Volumes of unit balls of mixed sequence spaces

The volume of the unit ball of the Lebesgue sequence space $\ell_p^m$ is very well known since the times of Dirichlet. We calculate the volume of the unit ball in the mixed norm $\ell^n_q(\ell_p^m)$, whose special cases are nowadays popular in machine learning under the name of group lasso. We consider the real as well as the complex case. The result is given by a closed formula involving the gamma function, only slightly more complicated than the one of Dirichlet. We close by an overview of open problems.

preprint2013arXiv

Lorentz spaces with variable exponents

We introduce Lorentz spaces $L_{p(\cdot),q}(\R^n)$ and $L_{p(\cdot),q(\cdot)}(\R^n)$ with variable exponents. We prove several basic properties of these spaces including embeddings and the identity $L_{p(\cdot),p(\cdot)}(\R^n)=L_{p(\cdot)}(\R^n)$. We also show that these spaces arise through real interpolation between $L_{\p}(\R^n)$ and $L_\infty(\R^n)$. Furthermore, we answer in a negative way the question posed in Diening, Hästö, and Nekvinda (2004) whether the Marcinkiewicz interpolation theorem holds in the frame of Lebesgue spaces with variable integrability.

preprint2012arXiv

Average best $m$-term approximation

We introduce the concept of average best $m$-term approximation widths with respect to a probability measure on the unit ball of $\ell_p^n$. We estimate these quantities for the embedding $id:\ell_p^n\to\ell_q^n$ with $0<p\le q\le \infty$ for the normalized cone and surface measure. Furthermore, we consider certain tensor product weights and show that a typical vector with respect to such a measure exhibits a strong compressible (i.e. nearly sparse) structure.

preprint2012arXiv

Non-smooth atomic decompositions, traces on Lipschitz domains, and pointwise multipliers in function spaces

We provide non-smooth atomic decompositions for Besov spaces $\Bd(\rn)$, $s>0$, $0<p,q\leq \infty$, defined via differences. The results are used to compute the trace of Besov spaces on the boundary $Γ$ of bounded Lipschitz domains $Ω$ with smoothness $s$ restricted to $0<s<1$ and no further restrictions on the parameters $p,q$. We conclude with some more applications in terms of pointwise multipliers.

preprint2011arXiv

Homogeneity property of Besov and Triebel-Lizorkin spaces

We consider the classical Besov and Triebel-Lizorkin spaces defined via differences and prove a homogeneity property for functions with bounded support in the frame of these spaces. As the proof is based on compact embeddings between the studied function spaces we present also some results on the entropy numbers of these embeddings. Moreover, we derive some applications in terms of pointwise multipliers.

preprint2010arXiv

A variant of the Johnson-Lindenstrauss lemma for circulant matrices

We continue our study of the Johnson-Lindenstrauss lemma and its connection to circulant matrices started in \cite{HV}. We reduce the bound on $k$ from $k=O(ε^{-2}\log^3n)$ proven there to $k=O(ε^{-2}\log^2n)$. Our technique differs essentially from the one used in \cite{HV}. We employ the discrete Fourier transform and singular value decomposition to deal with the dependency caused by the circulant structure.