Przemysław Wojtaszczyk
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Przemysław Wojtaszczyk has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- José Luis Ansorena
Universidad de La Rioja
10 shared publications - Fernando Albiac
Department of Mathematics, Statistics and Computer Sciences, and InaMat2, Universidad Pública de Navarra, Pamplona, Spain
9 shared publications - Guergana Petrova
Texas A&M University
7 shared publications - Pablo Manuel Berná
Departamento de Matemáticas, CUNEF Universidad, Madrid, Spain
2 shared publications - Albert Cohen
Sorbonne Universite
1 shared publication - G. L. Mullendore
Department of Atmospheric Sciences, University of North Dakota, Grand Forks, North Dakota, USA
1 shared publication
Latest Publications
- 29 June 2024
- Published by Springer Nature
- p. 19-29
Neural Networks: Deep, Shallow, or in Between?
- Open Access
Neural networks: deep, shallow, or in between?
We give estimates from below for the error of approximation of a compact subset from a Banach space by the outputs of feed-forward neural networks with width W, depth l and Lipschitz activation functions. We show that, modulo logarithmic factors, rates better that entropy numbers' rates are possibly attainable only for neural networks for which the depth l goes to infinity, and that there is no gain if we fix the depth and let the width W go to infinity.
Open AccessNeural networks: deep, shallow, or in between?
We give estimates from below for the error of approximation of a compact subset from a Banach space by the outputs of feed-forward neural networks with width W, depth l and Lipschitz activation functions. We show that, modulo logarithmic factors, rates better that entropy numbers' rates are possibly attainable only for neural networks for which the depth l goes to infinity, and that there is no gain if we fix the depth and let the width W go to infinity.
- Open Access
- 29 July 2023
- Published by Springer Nature in Constructive Approximation
- Vol. 60 (2) , 253-283
Existence of Almost Greedy Bases in Mixed-Norm Sequence and Matrix Spaces, Including Besov Spaces
We prove that the sequence spaces $$\ell _p\oplus \ell _q$$ and the spaces of infinite matrices $$\ell _p(\ell _q)$$ , $$\ell _q(\ell _p)$$ and $$(\bigoplus _{n=1}^\infty \ell _p^n)_{\ell _q}$$ , which are isomorphic to certain Besov spaces, have an almost greedy basis whenever $$0<p<1<q<\infty $$ . More precisely, we custom-build almost greedy bases in such a way that the Lebesgue parameters grow in a prescribed manner. Our arguments critically depend on the extension of the Dilworth–Kalton–Kutzarova method from Dilworth et al. (Stud Math 159(1):67–101, 2003), which was originally designed for constructing almost greedy bases in Banach spaces, to make it valid for direct sums of mixed-normed spaces with nonlocally convex components. Additionally, we prove that the fundamental functions of all almost greedy bases of these spaces grow as $$(m^{1/q})_{m=1}^\infty $$ .- 29 July 2023
- Published by Springer Nature in Constructive Approximation
- Vol. 60 (2) , 253-283
Open AccessExistence of Almost Greedy Bases in Mixed-Norm Sequence and Matrix Spaces, Including Besov Spaces
We prove that the sequence spaces $$\ell _p\oplus \ell _q$$ and the spaces of infinite matrices $$\ell _p(\ell _q)$$ , $$\ell _q(\ell _p)$$ and $$(\bigoplus _{n=1}^\infty \ell _p^n)_{\ell _q}$$ , which are isomorphic to certain Besov spaces, have an almost greedy basis whenever $$0<p<1<q<\infty $$ . More precisely, we custom-build almost greedy bases in such a way that the Lebesgue parameters grow in a prescribed manner. Our arguments critically depend on the extension of the Dilworth–Kalton–Kutzarova method from Dilworth et al. (Stud Math 159(1):67–101, 2003), which was originally designed for constructing almost greedy bases in Banach spaces, to make it valid for direct sums of mixed-normed spaces with nonlocally convex components. Additionally, we prove that the fundamental functions of all almost greedy bases of these spaces grow as $$(m^{1/q})_{m=1}^\infty $$ . - 27 December 2022
- Published by Springer Nature in Israel Journal of Mathematics
- Vol. 255 (2) , 621-644
Lorentz spaces and embeddings induced by almost greedy bases in superreflexive Banach spaces
- 27 December 2022
- Published by Springer Nature in Israel Journal of Mathematics
- Vol. 255 (2) , 621-644
Lorentz spaces and embeddings induced by almost greedy bases in superreflexive Banach spaces
- Open Access
Limitations on approximation by deep and shallow neural networks
We prove Carl's type inequalities for the error of approximation of compact sets K by deep and shallow neural networks. This in turn gives lower bounds on how well we can approximate the functions in K when requiring the approximants to come from outputs of such networks. Our results are obtained as a byproduct of the study of the recently introduced Lipschitz widths.
Open AccessLimitations on approximation by deep and shallow neural networks
We prove Carl's type inequalities for the error of approximation of compact sets K by deep and shallow neural networks. This in turn gives lower bounds on how well we can approximate the functions in K when requiring the approximants to come from outputs of such networks. Our results are obtained as a byproduct of the study of the recently introduced Lipschitz widths.
- 6 June 2022
- Published by Springer Nature in Constructive Approximation
- Vol. 57 (2) , 759-805
Lipschitz Widths
- 6 June 2022
- Published by Springer Nature in Constructive Approximation
- Vol. 57 (2) , 759-805
Lipschitz Widths
- Open Access
On the entropy numbers and the Kolmogorov widths
Direct estimates between linear or nonlinear Kolmogorov widths and entropy numbers are presented. These estimates are derived using the recently introduced Lipschitz widths. Applications for m-term approximation are obtained.
Open AccessOn the entropy numbers and the Kolmogorov widths
Direct estimates between linear or nonlinear Kolmogorov widths and entropy numbers are presented. These estimates are derived using the recently introduced Lipschitz widths. Applications for m-term approximation are obtained.
- 8 July 2021
- Published by Springer Nature in Foundations of Computational Mathematics
- Vol. 22 (3) , 607-648
Optimal Stable Nonlinear Approximation
- 8 July 2021
- Published by Springer Nature in Foundations of Computational Mathematics
- Vol. 22 (3) , 607-648
Optimal Stable Nonlinear Approximation
- 1 January 2021
- Published by Institute of Mathematics, Polish Academy of Sciences in Dissertationes Mathematicae
- Vol. 560, 1-88
Greedy approximation for biorthogonal systems in quasi-Banach spaces
The general problem addressed in this work is the development of a systematic study of the thresholding greedy algorithm for general biorthogonal systems %(also known as Markushevich bases) in quasi-Banach spaces from a functional-analytic point of v
- 1 January 2021
- Published by Institute of Mathematics, Polish Academy of Sciences in Dissertationes Mathematicae
- Vol. 560, 1-88
Greedy approximation for biorthogonal systems in quasi-Banach spaces
The general problem addressed in this work is the development of a systematic study of the thresholding greedy algorithm for general biorthogonal systems %(also known as Markushevich bases) in quasi-Banach spaces from a functional-analytic point of v
- 18 November 2020
- Published by Elsevier in Journal of Functional Analysis
- Vol. 280 (7) , 108871
Quasi-greedy bases in ℓ (0 < p < 1) are democratic
- 18 November 2020
- Published by Elsevier in Journal of Functional Analysis
- Vol. 280 (7) , 108871
Quasi-greedy bases in ℓ (0 < p < 1) are democratic