4.4 Article

Dual Wavelet Frames and Riesz Bases in Sobolev Spaces

Journal

CONSTRUCTIVE APPROXIMATION
Volume 29, Issue 3, Pages 369-406

Publisher

SPRINGER
DOI: 10.1007/s00365-008-9027-x

Keywords

Dual wavelet frames; Wavelet frames; Riesz bases; Sobolev spaces

Categories

Funding

  1. Alexander von Humboldt Foundation
  2. NSERC Canada [RGP 228051]
  3. National University of Singapore

Ask authors/readers for more resources

This paper generalizes the mixed extension principle in L-2(R-d) of (Ron and Shen in J. Fourier Anal. Appl. 3: 617-637, 1997) to a pair of dual Sobolev spaces H-s ( Rd) and H-s (R-d). In terms of masks for phi, psi(1),...,psi(L) is an element of H-s (R-d) and (phi) over tilde, (psi) over tilde (1),..., (psi) over tilde (L) is an element of H-s (R-d), simple sufficient conditions are given to ensure that (X-s(phi; psi(1),...,psi(L)), X-s ((phi) over tilde; (psi) over tilde (1),..., (psi) over tilde (L))) forms a pair of dual wavelet frames in (H-s(R-d), H-s (R-d)), where X-s (phi; psi(1),...,psi(L)) := {phi (.-k) : k is an element of Z(d)} boolean OR{2(j(d/2-s))psi(l)(2(j).-k) : j is an element of N-0, k is an element of Z(d), l = 1, ... , L}. For s > 0, the key of this general mixed extension principle is the regularity of phi, psi(1),..., psi(L), and the vanishing moments of (psi) over tilde (1),..., (psi) over tilde (L), while allowing (phi) over tilde, (psi) over tilde (1),..., (psi) over tilde (L) to be tempered distributions not in L-2(R-d) and psi(1),..., psi(L) to have no vanishing moments. So, the systems X-s(phi; psi(1),..., psi(L)) and X-s ((phi) over tilde; (psi) over tilde (1),..., (psi) over tilde (L)) may not be able to be normalized into a frame of L-2(R-d). As an example, we show that {2(j(1/2-s)) B-m(2(j).-k) : j is an element of N-0, k is an element of Z} is a wavelet frame in H-s (R) for any 0 < s < m-1/2, where B-m is the B-spline of order m. This simple construction is also applied to multivariate box splines to obtain wavelet frames with short supports, noting that it is hard to construct nonseparable multivariate wavelet frames with small supports. Applying this general mixed extension principle, we obtain and characterize dual Riesz bases (X-s(phi; psi(1),..., psi(L)), X-s ((phi) over tilde; (psi) over tilde (1),..., (psi) over tilde (L))) in Sobolev spaces (H-s(R-d), H-s (R-d)). For example, all interpolatory wavelet systems in (Donoho, Interpolating wavelet transform. Preprint, 1997) generated by an interpolatory refinable function phi is an element of H-s (R) with s > 1/2 are Riesz bases of the Sobolev space H-s (R). This general mixed extension principle also naturally leads to a characterization of the Sobolev norm of a function in terms of weighted norm of its wavelet coefficient sequence (decomposition sequence) without requiring that dual wavelet frames should be in L-2( R-d), which is quite different from other approaches in the literature.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.4
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

Article Mathematics, Applied

SYMMETRIC CANONICAL QUINCUNX TIGHT FRAMELETS WITH HIGH VANISHING MOMENTS AND SMOOTHNESS

Bin Han, Qingtang Jiang, Zuowei Shen, Xiaosheng Zhuang

MATHEMATICS OF COMPUTATION (2018)

Article Computer Science, Interdisciplinary Applications

B-spline tight frame based force matching method

Jianbin Yang, Guanhua Zhu, Dudu Tong, Lanyuan Lu, Zuowei Shen

JOURNAL OF COMPUTATIONAL PHYSICS (2018)

Article Mathematics, Interdisciplinary Applications

DIGITAL GABOR FILTERS WITH MRA STRUCTURE

Hut Ji, Zuowei Shen, Yufei Zhao

MULTISCALE MODELING & SIMULATION (2018)

Article Biochemistry & Molecular Biology

Investigating energy-based pool structure selection in the structure ensemble modeling with experimental distance constraints: The example from a multidomain protein Pub1

Guanhua Zhu, Wei Liu, Chenglong Bao, Dudu Tong, Hui Ji, Zuowei Shen, Daiwen Yang, Lanyuan Lu

PROTEINS-STRUCTURE FUNCTION AND BIOINFORMATICS (2018)

Article Computer Science, Artificial Intelligence

Coherence Retrieval Using Trace Regularization

Chenglong Bao, George Barbastathis, Hui Ji, Zuowei Shen, Zhengyun Zhang

SIAM JOURNAL ON IMAGING SCIENCES (2018)

Article Mathematics, Applied

Digital Gabor filters do generate MRA-based wavelet tight frames

Hui Ji, Zuowei Shen, Yufei Zhao

APPLIED AND COMPUTATIONAL HARMONIC ANALYSIS (2019)

Article Mathematics, Applied

Phase retrieval: A data-driven wavelet frame based approach

Tongyao Pang, Qingna Li, Zaiwen Wen, Zuowei Shen

APPLIED AND COMPUTATIONAL HARMONIC ANALYSIS (2020)

Article Computer Science, Artificial Intelligence

Nonlinear approximation via compositions

Zuowei Shen, Haizhao Yang, Shijun Zhang

NEURAL NETWORKS (2019)

Article Computer Science, Artificial Intelligence

Deep Network With Approximation Error Being Reciprocal of Width to Power of Square Root of Depth

Zuowei Shen, Haizhao Yang, Shijun Zhang

Summary: The new network, utilizing Floor-ReLU activation function, can approximate Holder functions and continuous functions under specific dual parameter conditions, overcoming the curse of dimensionality in approximation power.

NEURAL COMPUTATION (2021)

Article Mathematics, Applied

Optimal approximation rate of ReLU networks in terms of width and depth

Zuowei Shen, Haizhao Yang, Shijun Zhang

Summary: This paper focuses on the approximation ability of deep feed-forward neural networks in terms of width and depth, proving that ReLU networks with specific widths and depths can approximate Holder continuous functions with certain rates. The research results are optimal in terms of width and depth separately, showing better approximation performance compared to existing results.

JOURNAL DE MATHEMATIQUES PURES ET APPLIQUEES (2022)

Article Computer Science, Artificial Intelligence

Neural network approximation: Three hidden layers are enough

Zuowei Shen, Haizhao Yang, Shijun Zhang

Summary: This paper introduces a three-hidden-layer neural network with super approximation power, named Floor-Exponential-Step (FLES) networks, which can uniformly approximate Holder continuous functions with an exponential approximation rate. The network overcomes the curse of dimensionality in approximation power by utilizing specific activation functions and can extend the results to general bounded continuous functions.

NEURAL NETWORKS (2021)

Article Mathematics, Applied

APPROXIMATION FROM NOISY DATA

Bin Dong, Zuowei Shen, Jianbin Yang

Summary: This paper explores the approximation of functions from noisy and nonsmooth observed data, with a focus on sparse noise removal schemes. Theoretical analysis is presented, highlighting the importance of sparsity-based denoising for effective approximation. A new approximation scheme is proposed for large datasets to significantly reduce noise level and ensure asymptotic convergence.

SIAM JOURNAL ON NUMERICAL ANALYSIS (2021)

Article Mathematics, Applied

DEEP NETWORK APPROXIMATION FOR SMOOTH FUNCTIONS

Jianfeng Lu, Zuowei Shen, Haizhao Yang, Shijun Zhang

Summary: This paper establishes the optimal approximation error characterization of deep ReLU networks for smooth functions in terms of both width and depth simultaneously. Through local Taylor expansions and their deep ReLU network approximations, it shows that deep ReLU networks can approximate functions with a nearly optimal approximation error for arbitrary width and depth specified by N epsilon N+ and L epsilon N+, respectively. The estimate provided is nonasymptotic, valid for any chosen width and depth.

SIAM JOURNAL ON MATHEMATICAL ANALYSIS (2021)

Article Physics, Mathematical

Deep Network Approximation Characterized by Number of Neurons

Zuowei Shen, Haizhao Yang, Shijun Zhang

COMMUNICATIONS IN COMPUTATIONAL PHYSICS (2020)

Article Mathematics, Interdisciplinary Applications

MULTISCALE DISCRETE FRAMELET TRANSFORM FOR GRAPH-STRUCTURED SIGNALS

Hui Ji, Zuowei Shen, Yufei Zhao

MULTISCALE MODELING & SIMULATION (2020)

No Data Available