Remarks on Kneip's linear smoothers

Sören R. Künzel,David Pollard,Dana Yang

DOI: https://doi.org/10.48550/arXiv.1405.1744

2014-05-08

Abstract:We were trying to understand the analysis provided by Kneip (1994, Ordered Linear Smoothers). In particular we wanted to persuade ourselves that his results imply the oracle inequality stated by Tsybakov (2014, Lecture 8). This note contains our reworking of Kneip's ideas.

Statistics Theory

What problem does this paper attempt to address?

The Iyengar Inequality, Revisited

X. H. Wang,C. C. Xie,S. J. Yang

DOI: https://doi.org/10.1007/s10543-007-0153-9

2007-01-01

Abstract:Detailed analysis shows that the famous Iyengar inequality actually says that the Trapezoidal formula is a central algorithm for approximating integrals over an appropriate interval for the class of functions whose derivatives are bounded by a positive number K in L ∞ -sense. The inherent nonlinearity from central algorithms reflects the importance of the Iyengar inequality and thus makes familiar linear methods malfunction when one tries to generalize it. It is shown that the generalization depends on a nonlinear system of equations satisfied by a set of free nodes of a perfect spline. Explicit constructions are obtained in the spirit of the Iyengar inequality for the class of functions whose r th ( r ≤4) derivatives are bounded by a positive number K in L ∞ -sense because a closed solution to the nonlinear system is only available for r ≤4. Connections with computational mathematics, especially with best interpolation and best quadrature, are discussed. Numerical experiments are also included.
A New Proof of Some Polynomial Inequalities Related to Pseudo-Splines

Qun Mo,Yi Shen,Song Li

DOI: https://doi.org/10.1016/j.acha.2007.07.002

IF: 2.974

2007-01-01

Applied and Computational Harmonic Analysis

Abstract:Pseudo-splines of type I were introduced in [I. Daubechies, B. Han, A. Ron, Z. Shen, Framelets: MRA-based constructions of wavelet frames, Appl. Comput. Harmon. Anal. 14 (2003) 1–46] and [Selenick, Smooth wavelet tight frames with zero moments, Appl. Comput. Harmon. Anal. 10 (2000) 163–181] and type II were introduced in [B. Dong, Z. Shen, Pseudo-splines, wavelets and framelets, Appl. Comput. Harmon. Anal. 22 (2007) 78–104]. Both types of pseudo-splines provide a rich family of refinable functions with B-splines, interpolatory refinable functions and refinable functions with orthonormal shifts as special examples. In [B. Dong, Z. Shen, Pseudo-splines, wavelets and framelets, Appl. Comput. Harmon. Anal. 22 (2007) 78–104], Dong and Shen gave a regularity analysis of pseudo-splines of both types. The key to regularity analysis is Proposition 3.2 in [B. Dong, Z. Shen, Pseudo-splines, wavelets and framelets, Appl. Comput. Harmon. Anal. 22 (2007) 78–104], which also appeared in [A. Cohen, J.P. Conze, Régularité des bases d'ondelettes et mesures ergodiques, Rev. Mat. Iberoamericana 8 (1992) 351–365] and [I. Daubechies, Ten Lectures on Wavelets, CBMS-NSF Series in Applied Mathematics, SIAM, Philadelphia, 1992] for the case l=N−1. In this note, we will give a new insight into this proposition.
Smoothed Analysis of the Komlós Conjecture: Rademacher Noise

Elad Aigner-Horev,Dan Hefetz,Michael Trushkin

2024-07-04

Abstract:The {\em discrepancy} of a matrix $M \in \mathbb{R}^{d \times n}$ is given by $\mathrm{DISC}(M) := \min_{\boldsymbol{x} \in \{-1,1\}^n} \|M\boldsymbol{x}\|_\infty$. An outstanding conjecture, attributed to Komlós, stipulates that $\mathrm{DISC}(M) = O(1)$, whenever $M$ is a Komlós matrix, that is, whenever every column of $M$ lies within the unit sphere. Our main result asserts that $\mathrm{DISC}(M + R/\sqrt{d}) = O(d^{-1/2})$ holds asymptotically almost surely, whenever $M \in \mathbb{R}^{d \times n}$ is Komlós, $R \in \mathbb{R}^{d \times n}$ is a Rademacher random matrix, $d = \omega(1)$, and $n = \omega(d \log d)$. The factor $d^{-1/2}$ normalising $R$ is essentially best possible and the dependency between $n$ and $d$ is asymptotically best possible. Our main source of inspiration is a result by Bansal, Jiang, Meka, Singla, and Sinha (ICALP 2022). They obtained an assertion similar to the one above in the case that the smoothing matrix is Gaussian. They asked whether their result can be attained with the optimal dependency $n = \omega(d \log d)$ in the case of Bernoulli random noise or any other types of discretely distributed noise; the latter types being more conducive for Smoothed Analysis in other discrepancy settings such as the Beck-Fiala problem. For Bernoulli noise, their method works if $n = \omega(d^2)$. In the case of Rademacher noise, we answer the question posed by Bansal, Jiang, Meka, Singla, and Sinha. Our proof builds upon their approach in a strong way and provides a discrete version of the latter. Breaking the $n = \omega(d^2)$ barrier and reaching the optimal dependency $n = \omega(d \log d)$ for Rademacher noise requires additional ideas expressed through a rather meticulous counting argument, incurred by the need to maintain a high level of precision all throughout the discretisation process.

Combinatorics,Discrete Mathematics,Information Theory,Probability
Elementary Proofs for Kato Smoothing Estimates of Schrödinger-like Dispersive Equations

Xuwen Chen

DOI: https://doi.org/10.1090/conm/581/11487

2010-07-13

Abstract:In this expository note, we prove some extensions and refinements of classical Kato type estimates with elementary techniques.

Analysis of PDEs
Quantitative Analysis for Perturbed Abstract Inequality Systems in Banach Spaces

Chong Li,K. F. Ng

DOI: https://doi.org/10.1137/17m1151274

IF: 2.763

2018-01-01

SIAM Journal on Optimization

Abstract:Using the error bound results established in the present paper for approximate solutions, we study the stability issues when perturbed by possibly nonaffine smooth maps E for the abstract inequality system F >=(K) 0 defined by a (possible nonclosed) convex cone K and a Frechet differentiable function F satisfying the (extended) weak gamma-condition. We provide some sufficient conditions, in terms of the information at a solution x(0), for ensuring the lower semicontinuity and/or the Lipschitz-like continuity at x(0) of the solution mapping for the perturbed system F + E >=(K) 0 with smooth perturbation E. Explicit upper bounds of the Lipschitz-like moduli are also provided.
Smoothed analysis of algorithms

Daniel A. Spielman,Shang-Hua Teng

DOI: https://doi.org/10.48550/arXiv.math/0212413

2002-12-01

Abstract:Spielman and Teng introduced the smoothed analysis of algorithms to provide a framework in which one could explain the success in practice of algorithms and heuristics that could not be understood through the traditional worst-case and average-case analyses. In this talk, we survey some of the smoothed analyses that have been performed.

Optimization and Control
A unified approach to inequalities for K-functionals and moduli of smoothness

Amiran Gogatishvili,Bohumír Opic,Sergey Tikhonov,Walter Trebels

DOI: https://doi.org/10.1007/s00209-024-03484-x

2024-04-30

Mathematische Zeitschrift

Abstract:The paper provides a detailed study of crucial inequalities for smoothness and interpolation characteristics in rearrangement invariant Banach function spaces. We present a unified approach based on Holmstedt formulas to obtain these estimates. As examples, we derive new inequalities for moduli of smoothness and K -functionals in various Lorentz spaces.

mathematics
Anisotropic refinable functions and the tile B-splines

Vladimir Yu. Protasov,Tatyana Zaitseva

2023-12-18

Abstract:The regularity of refinable functions has been analysed in an extensive literature and is well-understood in two cases: 1) univariate 2) multivariate with an isotropic dilation matrix. The general (non-isotropic) case offered a great resistance. It was done only recently by developing the matrix method. In this paper we make the next step and extend the Littlewood-Paley type method, which is very efficient in the aforementioned special cases, to general equations with arbitrary dilation matrices. This gives formulas for the higher order regularity in $W_2^k(\mathbb{R}^n)$ by means of the Perron eigenvalue of a finite-dimensional linear operator on a special cone. Applying those results to recently introduced tile B-splines, we prove that they can have a higher smoothness than the classical ones of the same order. Moreover, the two-digit tile B-splines have the minimal support of the mask among all refinable functions of the same order of approximation. This proves, in particular, the lowest algorithmic complexity of the corresponding subdivision schemes. Examples and numerical results are provided.

Functional Analysis,Numerical Analysis
Inhomogeneous Hopf-Ole\u{i}nik Lemma and Applications. Part IV: Sharp Krylov Boundary Gradient Type Estimates for Solutions to Fully Nonlinear Differential Inequalities with unbounded coefficients and $C^{1,Dini}$ boundary data

J. Ederson M. Braga,Diego Moreira,Lihe Wang

2016-01-01

Abstract: In this paper we provide another application of the Inhomogeneous Hopf-Ole\u{i}nik Lemma (IHOL) proved in \cite{BM-IHOL-PartI}. As a matter of fact, we also provide a new and simpler proof of IHOL with unbounded coefficients but only for the the fully nonlinear differential inequalities case. The paper has essentially two parts. In the first part, we use IHOL for unbounded RHS to develop a Caffarelli's "Lipschitz implies $C^{1,\alpha}$" approach to prove Ladyzhenskaya-Uraltseva boundary gradient type estimates for functions in $S^{*}(\gamma, f)$ that vanishes on the boundary. Here, unbounded RHS means that $f\in L^{q}$ with $q>n$. This extends the celebrated Krylov's boundary gradient estimate proved in \cite{Krylov}. A Phragm\'en-Lindel\"of classification result for solutions in half spaces is recovered from these estimates. In the second part, we extend the previous results for functions in $S^{*}(\gamma, \sigma, f)$ where $\gamma,f\in L^{q}$ with $q>n$ that have a $C^{1,Dini}$ boundary data on a $W^{2,q}$ domain. Here, we use an "improvement of flatness" strategy suited to the unbounded coefficients scenario. As a consequence of that, a quantitative version of IHOL under pointwise $C^{1,Dini}$ boundary regularity is obtained.
On the refined Strichartz estimates

Ciprian Demeter

DOI: https://doi.org/10.48550/arXiv.2002.09525

2020-02-22

Abstract:We present a slightly simpler proof of the multilinear refined Strichartz estimate, and prove a slightly more general linear refined Strichartz estimate. Our arguments seek to clarify the connection between these estimates, refined decoupling and tube incidences.

Classical Analysis and ODEs,Analysis of PDEs
Another Refinement of the Pólya–Szegö Inequality

Yu-Dong Wu,V. Lokesha,H. M. Srivastava

DOI: https://doi.org/10.1016/j.camwa.2010.05.023

2010-01-01

Abstract:In this paper, the authors make use of certain analytical techniques for nonlinear algebraic equation systems in order to give another refinement of the Pólya–Szegö inequality in a triangle, which is associated with one of Chen’s theorems (see Chen (1993) [12] and Chen (2000) [13]). Some remarks and observations, as well as two closely-related open problems, are also presented.
Several inequalities concerning interpolation in classical Fourier analysis

Runzhe Zhang,Hua Wang

2023-03-02

Abstract:In this note, we establish several interpolation inequalities in $\mathbb R^n$ in the Lebesgue spaces and Morrey spaces. By using the classical Calderon--Zygmund decomposition, we will reprove that $L^{p}(\mathbb R^n)\cap\mathrm{BMO}(\mathbb R^n)\subset L^{q}(\mathbb R^n)$ for all $q$ with $p<q<\infty$, where $1\leq p<\infty$. We also reprove that there exists a constant $C(p,q,n)$ depending on $p,q,n$ such that the following inequality \begin{equation*} \|f\|_{L^q}\leq C(p,q,n)\cdot\big(\|f\|_{L^p}\big)^{p/q}\cdot\big(\|f\|_{\mathrm{BMO}}\big)^{1-p/q} \end{equation*} holds for all $f\in L^{p}(\mathbb R^n)\cap\mathrm{BMO}(\mathbb R^n)$ with $1\leq p<\infty$. Moreover, this embedding constant has the optimal growth order $q$ as $q\to\infty$, which was given by Chen--Zhu, and Kozono--Wadade. We will show that $L^{p,\kappa}(\mathbb R^n)\cap\mathrm{BMO}(\mathbb R^n)\subset L^{q,\kappa}(\mathbb R^n)$ for all $q$ with $p<q<\infty$, where $1\leq p<\infty$ and $0<\kappa<1$. Moreover, there exists a constant $\widetilde{C}(p,q,n)$ depending on $p,q,n$ such that \begin{equation*} \|f\|_{L^{q,\kappa}}\leq \widetilde{C}(p,q,n)\cdot\big(\|f\|_{L^{p,\kappa}}\big)^{p/q}\cdot\big(\|f\|_{\mathrm{BMO}}\big)^{1-p/q} \end{equation*} holds for all $f\in L^{p,\kappa}(\mathbb R^n)\cap\mathrm{BMO}(\mathbb R^n)$ with $1\leq p<\infty$ and $0<\kappa<1$. This embedding constant is shown to have the linear growth order as $q\to\infty$, that is, $\widetilde{C}(p,q,n)\leq C_n\cdot q$ with the constant $C_n$ depending only on the dimension $n$, when $q$ is large. As an application of the above results, some new bilinear estimates are also established, which can be used in the study of the global existence and regularity of weak solutions to elliptic and parabolic partial differential equations of the second order.

Classical Analysis and ODEs
Remarks on Large-Scale Effects of Smoothing Mechanisms in 3D Reaction-Diffusion Equations

Dirk Erhard,Weijun Xu

2021-01-01

Markov processes and related fields

Abstract:In [2], we considered a class of reaction-diffusion equations that approximates the dynamical Phi(4)(3) model at large scales. These approximations involve very general smoothing mechanisms that are higher order perturbations of the Laplacian. In this note, we discuss assumptions made on the smoothing mechanism in that article. In particular, we remark that the strict positivity condition of the Fourier multiplier of the operator cannot be relaxed without other modifications, not even to allow it to reach zero outside the origin. On the other hand, if we introduce suitable Fourier cutoffs that are compatible with the non-positive part of the operator, then we expect that no smoothing assumption will be needed on its higher order terms. We then explain how this changes the coupling constant of the limiting equation, and how to modify the argument to prove the result.
The Skein Polynomial for Links

Boju Jiang,Jiajun Wang,Hao Zheng

DOI: https://doi.org/10.1142/s0218216517420032

2017-01-01

Journal of Knot Theory and Its Ramifications

Abstract:We give characterizations of the skein polynomial for links (as well as Jones and Alexander-Conway polynomials derivable from it), avoiding the usual "smoothing of a crossing" move. As by-products we have characterizations of these polynomials for knots, and for links with any given number of components.
Small noise analysis for Tikhonov and RKHS regularizations

Quanjun Lang,Fei Lu

2024-09-04

Abstract:Regularization plays a pivotal role in ill-posed machine learning and inverse problems. However, the fundamental comparative analysis of various regularization norms remains open. We establish a small noise analysis framework to assess the effects of norms in Tikhonov and RKHS regularizations, in the context of ill-posed linear inverse problems with Gaussian noise. This framework studies the convergence rates of regularized estimators in the small noise limit and reveals the potential instability of the conventional L2-regularizer. We solve such instability by proposing an innovative class of adaptive fractional RKHS regularizers, which covers the L2 Tikhonov and RKHS regularizations by adjusting the fractional smoothness parameter. A surprising insight is that over-smoothing via these fractional RKHSs consistently yields optimal convergence rates, but the optimal hyper-parameter may decay too fast to be selected in practice.

Machine Learning
Upper and Lower Bounds on the Smoothed Complexity of the Simplex Method

Sophie Huiberts,Yin Tat Lee,Xinzhi Zhang

2024-05-16

Abstract:The simplex method for linear programming is known to be highly efficient in practice, and understanding its performance from a theoretical perspective is an active research topic. The framework of smoothed analysis, first introduced by Spielman and Teng (JACM '04) for this purpose, defines the smoothed complexity of solving a linear program with $d$ variables and $n$ constraints as the expected running time when Gaussian noise of variance $\sigma^2$ is added to the LP data. We prove that the smoothed complexity of the simplex method is $O(\sigma^{-3/2} d^{13/4}\log^{7/4} n)$, improving the dependence on $1/\sigma$ compared to the previous bound of $O(\sigma^{-2} d^2\sqrt{\log n})$. We accomplish this through a new analysis of the \emph{shadow bound}, key to earlier analyses as well. Illustrating the power of our new method, we use our method to prove a nearly tight upper bound on the smoothed complexity of two-dimensional polygons.

Data Structures and Algorithms
Gagliardo-Nirenberg inequality via a new pointwise estimate

Karol Lesnik,Tomas Roskovec,Filip Soudsky

2024-03-12

Abstract:We prove a new type of pointwise estimate of the Kalamajska-Mazya-Shaposhnikova type, where sparse averaging operators replace the maximal operator. It allows us to extend the Gagliardo-Nirenberg interpolation inequality to all rearrangement invariant Banach function spaces without any assumptions on their upper Boyd index, i.e. omitting problems caused by unboundedness of maximal operator on spaces close to $L^1$. In particular, we remove unnecessary assumptions from the Gagliardo-Nirenberg inequality in the setting of Orlicz and Lorentz spaces. The applied method is new in this context and may be seen as a kind of sparse domination technique fitted to the context of rearrangement invariant Banach function spaces.

Functional Analysis,Analysis of PDEs
Error Bounds for a Kernel-Based Constrained Optimal Smoothing Approximation

Laurence Grammont,François Bachoc,Andrés F. López-Lopera

2024-07-12

Abstract:This paper establishes error bounds for the convergence of a piecewise linear approximation of the constrained optimal smoothing problem posed in a reproducing kernel Hilbert space (RKHS). This problem can be reformulated as a Bayesian estimation problem involving a Gaussian process related to the kernel of the RKHS. Consequently, error bounds can be interpreted as a quantification of the maximum a posteriori (MAP) accuracy. To our knowledge, no error bounds have been proposed for this type of problem so far. The convergence results are provided as a function of the grid size, the regularity of the kernel, and the distance from the kernel interpolant of the approximation to the set of constraints. Inspired by the MaxMod algorithm from recent literature, which sequentially allocates knots for the piecewise linear approximation, we conduct our analysis for non-equispaced knots. These knots are even allowed to be non-dense, which impacts the definition of the optimal smoothing solution and our error bound quantifiers. Finally, we illustrate our theorems through several numerical experiments involving constraints such as boundedness and monotonicity.

Statistics Theory
Free knot linear interpolation and the Milstein scheme for stochastic differential equations

Mehdi Slassi

DOI: https://doi.org/10.48550/arXiv.1306.4457

2013-06-19

Probability

Abstract:The main purpose of this paper is to give a solution to a long-standing unsolved problem concerning the pathwise strong approximation of stochastic differential equations with respect to the global error in the $L_{\infty}$-norm. Typically, one has average supnorm error of order $\left(\ln k/k\right)^{1/2}$ for standard approximations of SDEs with $k$ discretization points, like piecewise interpolated Ito-Taylor schemes. On the other hand there is a lower bound, which indicates that the order $1/\sqrt{k}$ is best possible for spline approximation of SDEs with $k$ free knots. The present paper deals with the question of how to get an implementable method, which achieves the order $1/\sqrt{k}$. Up to now, papers with regard to this issue give only pure existence results and so are inappropriate for practical use. In this paper we introduce a nonlinear method for approximating a scalar SDE. The method combines a Milstein scheme with a pieceweise linear interpolation of the Brownian motion with free knots and is easy to implement. Moreover, we establish sharp lower and upper error bounds with specified constants which exhibit the influence of the coefficients of the equation.

Remarks on Kneip's linear smoothers

The Iyengar Inequality, Revisited

A New Proof of Some Polynomial Inequalities Related to Pseudo-Splines

Smoothed Analysis of the Komlós Conjecture: Rademacher Noise

Elementary Proofs for Kato Smoothing Estimates of Schrödinger-like Dispersive Equations

Quantitative Analysis for Perturbed Abstract Inequality Systems in Banach Spaces

Smoothed analysis of algorithms

A unified approach to inequalities for K-functionals and moduli of smoothness

Anisotropic refinable functions and the tile B-splines

Inhomogeneous Hopf-Ole\u{i}nik Lemma and Applications. Part IV: Sharp Krylov Boundary Gradient Type Estimates for Solutions to Fully Nonlinear Differential Inequalities with unbounded coefficients and $C^{1,Dini}$ boundary data

On the refined Strichartz estimates

Another Refinement of the Pólya–Szegö Inequality

Several inequalities concerning interpolation in classical Fourier analysis

Remarks on Large-Scale Effects of Smoothing Mechanisms in 3D Reaction-Diffusion Equations

The Skein Polynomial for Links

Small noise analysis for Tikhonov and RKHS regularizations

Upper and Lower Bounds on the Smoothed Complexity of the Simplex Method

Gagliardo-Nirenberg inequality via a new pointwise estimate

Error Bounds for a Kernel-Based Constrained Optimal Smoothing Approximation

Free knot linear interpolation and the Milstein scheme for stochastic differential equations