Different Degrees Of Smoothness Research Articles

An adaptive functional regression framework for locally heterogeneous signals in spectroscopy

Abstract In recent years, there has been growing attention towards food nutritional properties, traceability, and production systems prioritizing environmental sustainability. Consequently, there is a rising demand for tools evaluating food quality and authenticity, with mid-infrared (MIR) spectroscopy techniques playing a pivotal role to collect vast amounts of data. These data pose some challenges that existing methods struggle to address, thus necessitating the development of new statistical techniques. We introduce an adaptive functional regression framework allowing for the definition of a flexible estimator accommodating different degrees of smoothness. We provide an optimization procedure handling both Gaussian and non-Gaussian responses, and allowing for the inclusion of scalar covariates. Our proposal is applied to MIR spectroscopy data, providing excellent performances when predicting milk composition and cows’ dietary regimens. Furthermore, the developed inferential routine enhances the interpretability of the results, providing valuable insights leading to a deeper understanding of the relation between specific wavenumbers and milk characteristics.

A flexible modulated pesticide release platform through poly(urethane-urea) microcapsules: effect of different crosslinkers compositions.

Polymeric microcapsules (MCs) have become an important issue and have attracted increasing attention because of their tunable physical and chemical properties. Diverse shell structures can confer multiple properties on MCs. Different polyols (1,4-butanediol and glycerin) and polyamines (triethylenetetramine and isophorondiamine) were selected as crosslinkers to obtain emamectin benzoate (EB)-loaded poly(urethane-urea) MCs (PU-MCs) by interfacial polymerization. The four obtained PU-MCs showed sphericity with different degrees of smoothness on their surfaces, and displayed a uniform size distribution ranging from 500 to 700 nm. Moreover, transmission electron microscopy showed that the shell thickness was roughly uniform, and was greatly influenced by the type and structure of the crosslinker. GI-MCs, prepared using glycerin and isophorondiamine, had the largest shell thickness. GT-MCs, obtained using glycerin and triethylenetetramine, had the highest encapsulation efficiency and drug-loading content, and BT-MCs, obtained using mixtures of 1,4-butanediol and triethylenetetramine, had the fastest release behavior. Thermogravimetric analysis revealed that the greater the degree of shell crosslinking, the higher decomposition temperature and the greater the thermal stability. A BT-MC suspension had the lowest viscosity and contact angle with the best wettability. Bioassay experiments showed that BT-MCs exhibited good insecticidal activity against Plutella xylostella larvae with a half-maximal lethal concentration of 4.19 mg/L. Furthermore, a BT-MC suspension showed good thermal and light stability, with potential applications in minimizing the toxicity of EB through sustained release. Various properties of EB-loaded PU-MCs were modulated through simple selection of different polyols and polyamines during fabrication, which might have an important role in constructing the pesticide delivery system and improving pesticide utilization. © 2024 Society of Chemical Industry.

A generalization of Floater–Hormann interpolants

In this paper the interpolating rational functions introduced by Floater and Hormann are generalized leading to a whole new family of rational functions depending on γ, an additional positive integer parameter. For γ=1, the original Floater–Hormann interpolants are obtained. When γ>1 we prove that the new rational functions share a lot of the nice properties of the original Floater–Hormann functions. Indeed, for any configuration of nodes in a compact interval, they have no real poles, interpolate the given data, preserve the polynomials up to a certain fixed degree, and have a barycentric-type representation. Moreover, we estimate the associated Lebesgue constants in terms of the minimum (h∗) and maximum (h) distance between two consecutive nodes. It turns out that, in contrast to the original Floater–Hormann interpolants, for all γ>1 we get uniformly bounded Lebesgue constants in the case of equidistant and quasi-equidistant nodes configurations (i.e., when h∼h∗). For such configurations, as the number of nodes tends to infinity, we prove that the new interpolants (γ>1) uniformly converge to the interpolated function f, for any continuous function f and all γ>1. The same is not ensured by the original FH interpolants (γ=1). Moreover, we provide uniform and pointwise estimates of the approximation error for functions having different degrees of smoothness. Numerical experiments illustrate the theoretical results and show a better error profile for less smooth functions compared to the original Floater–Hormann interpolants.

Bi-smooth constraints for accelerated dynamic MRI with low-rank plus sparse tensor decomposition

The long data acquisition required for dynamic magnetic resonance imaging (DMRI) has limited its development. This paper proposes a low-rank plus sparse tensor decomposition model with bi-smooth constraints to reconstruct DMR images from under-sampled k-t space data. The DMR images can be naturally decomposed into a low-rank and sparse tensor to exploit the spatiotemporal correlation to improve the reconstruction quality. Different degrees of smoothness in the foreground and background parts of the DMR images are analyzed and utilized to further exploit the smoothness with a divide-and-conquer strategy. By imposing the bi-smooth constraints on these two parts, more edges and details of DMR images can be preserved. The experimental results show that the proposed method can achieve better reconstruction performance than the existing state-of-the-art methods.

VOLDIS: A Direct Search for the Size Distribution of Nanoparticles from Small-Angle Scattering Data and Stability of the Solution

The paper considers an algorithm for the direct search for a nonparametric smooth histogram of the particle size distribution from small-angle X-ray and neutron scattering data. The features and details of the implementation of the method, which consists in the sequential search for several solutions with different degrees of smoothness of the distribution contour, are considered. Methods for evaluating the stability of both the whole distribution contour and its individual parts are discussed. The work of the program is illustrated by examples of the analysis of polydisperse spherical particles in silicasol solutions.

Wavelet-based robust estimation and variable selection in nonparametric additive models

This article studies M-type estimators for fitting robust additive models in the presence of anomalous data. The components in the additive model are allowed to have different degrees of smoothness. We introduce a new class of wavelet-based robust M-type estimators for performing simultaneous additive component estimation and variable selection in such inhomogeneous additive models. Each additive component is approximated by a truncated series expansion of wavelet bases, making it feasible to apply the method to nonequispaced data and sample sizes that are not necessarily a power of 2. Sparsity of the additive components together with sparsity of the wavelet coefficients within each component (group), results into a bi-level group variable selection problem. In this framework, we discuss robust estimation and variable selection. A two-stage computational algorithm, consisting of a fast accelerated proximal gradient algorithm of coordinate descend type, and thresholding, is proposed. When using nonconvex redescending loss functions, and appropriate nonconvex penalty functions at the group level, we establish optimal convergence rates of the estimates. We prove variable selection consistency under a weak compatibility condition for sparse additive models. The theoretical results are complemented with some simulations and real data analysis, as well as a comparison to other existing methods.

A cost-effective algorithm for calibrating multiscale geographically weighted regression models

ABSTRACT The multiscale geographically weighted regression (MGWR) model is a useful extension of the geographically weighted regression (GWR) model. MGWR, however, is a kind of Nadaraya–Watson kernel smoother, which usually leads to inaccurate estimates for the regression function and suffers from the boundary effect. Moreover, the widely used calibration technique for the MGWR with a back-fitting estimator (MGWR-BF) is computationally demanding, preventing it from being applied to large-scale data. To overcome these problems, we proposed a local linear-fitting-based MGWR (MGWR-LL) by introducing a local spatially varying coefficient model in which coefficients of different variables could be characterised as linear functions of spatial coordinates with different degrees of smoothness. Then the model was calibrated with a two-step least-squared estimated algorithm. Both simulated and actual data were implemented to validate the performance of the proposed method. The results consistently showed that the MGWR-LL automatically corrected for the boundary effect and improved the accuracy in most cases, not only in the goodness-of-fit measure but also in reducing the bias of the coefficient estimates. Moreover, the MGWR-LL significantly outperformed the MGWR-BF in computational cost, especially for larger-scale data. These results demonstrated that the proposed method can be a useful tool for the MGWR calibration.

Generalized A-GPSFS and modeling of vibrating elements with internal complexities

Mechanical components containing internal complexities in free vibration mode are analyzed through generalized bases made of global piecewise smooth functions. Such bases introduced in 2002, as GPSFs, were made adaptive in 2015 to reduce the computational efforts (A-GPSFs) while, a recent study in 2018 used supplementary analytical conditions along with A-GPSFs to study free vibrations of thin-walled beams and plates. In this paper, generalized bases of A-GPSFs are introduced (i) in a self-contained manner, id est, without needing any supplementary analytical conditions and even (ii) extending the possibility of modeling mechanical components containing internal jumps. It is shown how such generalized A-GPSFs can be used to model continuous functions which have different degrees of smoothness; even discontinuous functions (i.e. containing jumps) can be modeled and always without involving significant approximations or letting rise to any Gibbs phenomenon. Thus, a free vibration analysis of mechanical components including internal complexities of engineering interest becomes part of a simple unified procedure based on a global approximation where unknown eigenvalues and eigenfunctions can be singled out only through generalized bases of A-GPSFs; the results of previous formulations are then obtained as particular cases. The efficiency and ability of the proposed models result from the comparison between calculated eigen-parameters and those of other models presented in the open literature.

New efficient spline estimation for varying-coefficient models with two-step knot number selection

One of the advantages for the varying-coefficient model is to allow the coefficients to vary as smooth functions of other variables and the coefficients functions can be estimated easily through a simple B-spline approximations method. This leads to a simple one-step estimation procedure. We show that such a one-step method cannot be optimal when some coefficient functions possess different degrees of smoothness. Under the regularity conditions, the consistency and asymptotic normality of the two step B-spline estimators are also derived. A few simulation studies show that the gain by the two-step procedure can be quite substantial. The methodology is illustrated by an AIDS data set.

Force balance in convectively driven dynamos with no inertia

We study dynamo action in rotating, plane layer Boussinesq convection in the absence of inertia. This allows a decomposition of the velocity into a thermal part driven by buoyancy, and a magnetic part driven by the Lorentz force. We have identified three families of solutions, defined in terms of what is the dominant contribution to the velocity. In weak field dynamos the dominant contribution is the thermal component, in super strong field dynamos the dominant contribution is magnetic and in strong field dynamos the two components are comparable. For each of these solutions we investigate the force balance in the momentum equation to determine the relative importance of the viscous, buoyancy, Coriolis and magnetic forces. We do this by extracting the solenoidal part of the individual terms in the momentum equation, thereby removing their pressure contributions. This is numerically preferable to the more common practice of taking the curl of the momentum equation, which introduces an extra derivative. We find that, irrespective of the type of dynamo solution, the dynamics is controlled by the horizontal forces (in projection). Furthermore, in the progression from weak to strong to super strong dynamos, we find that the viscous forces in the thermal equation become negligible, thereby leading to a balance between buoyancy and Coriolis forces. On the other hand, no corresponding trend is observed in the magnetic part of the momentum equation: the viscous stresses always remain significant. This can be attributed to the different degrees of smoothness of the Coriolis and Lorentz forces, the latter having contributions from strong, filamentary structures. We discuss how our findings relate to dynamo solutions in which viscosity plays no role whatsoever – so-called Taylor states.

Single image super-resolution based on approximated Heaviside functions and iterative refinement.

One method of solving the single-image super-resolution problem is to use Heaviside functions. This has been done previously by making a binary classification of image components as “smooth” and “non-smooth”, describing these with approximated Heaviside functions (AHFs), and iteration including l1 regularization. We now introduce a new method in which the binary classification of image components is extended to different degrees of smoothness and non-smoothness, these components being represented by various classes of AHFs. Taking into account the sparsity of the non-smooth components, their coefficients are l1 regularized. In addition, to pick up more image details, the new method uses an iterative refinement for the residuals between the original low-resolution input and the downsampled resulting image. Experimental results showed that the new method is superior to the original AHF method and to four other published methods.

An alternative bandwidth selection method for estimating functional coefficient models

Functional coefficient regression models are very useful for many statistics and economics applications and there exists a large body of literature on kernel estimation of the coefficient functions. Fan and Zhang (1999) point out that the traditional estimation method is not optimal if the coefficient functions possess different degrees of smoothness. They propose a two-step method to attenuate the drawback. To apply the two-step method, one needs to identify which coefficient function is the smoothest one and undersmooth all other functions in the first stage. In this paper we propose an alternative approach which assigns each smooth function a different bandwidth and we choose the bandwidths simultaneously. Simulation results show that our proposed method compare well with existing methods.

Time-varying coefficients models for recurrent event data when different varying coefficients admit different degrees of smoothness: application to heart disease modeling.

We consider a class of semiparametric marginal rate models for analyzing recurrent event data. In these models, both time-varying and time-free effects are present, and the estimation of time-varying effects may result in non-smooth regression functions. A typical approach for avoiding this problem and producing smooth functions is based on kernel methods. The traditional kernel-based approach, however, assumes a common degree of smoothness for all time-varying regression functions, which may result in suboptimal estimators if the functions have different levels of smoothness. In this paper, we extend the traditional approach by introducing different bandwidths for different regression functions. First, we establish the asymptotic properties of the suggested estimators. Next, we demonstrate the superiority of our proposed method using two finite-sample simulation studies. Finally, we illustrate our methodology by analyzing a real-world heart disease dataset. Copyright © 2016 John Wiley & Sons, Ltd.

Integral approximation by kernel smoothing

Let $(X_1,\ldots,X_n)$ be an i.i.d. sequence of random variables in $\mathbb{R}^d$, $d\geq 1$. We show that, for any function $\varphi :\mathbb{R}^d\rightarrow\mathbb{R}$, under regularity conditions, \[n^ {1/2}\Biggl(n^{-1}\sum_{i=1}^n\frac{\varphi(X_i)}{\widehat{f}^(X_i)}- \int \varphi(x)\,dx\Biggr)\stackrel{\mathbb{P}}{\longrightarrow}0,\] where $\widehat{f}$ is the classical kernel estimator of the density of $X_1$. This result is striking because it speeds up traditional rates, in root $n$, derived from the central limit theorem when $\widehat{f}=f$. Although this paper highlights some applications, we mainly address theoretical issues related to the later result. We derive upper bounds for the rate of convergence in probability. These bounds depend on the regularity of the functions $\varphi$ and $f$, the dimension $d$ and the bandwidth of the kernel estimator $\widehat{f}$. Moreover, they are shown to be accurate since they are used as renormalizing sequences in two central limit theorems each reflecting different degrees of smoothness of $\varphi$. As an application to regression modelling with random design, we provide the asymptotic normality of the estimation of the linear functionals of a regression function. As a consequence of the above result, the asymptotic variance does not depend on the regression function. Finally, we debate the choice of the bandwidth for integral approximation and we highlight the good behavior of our procedure through simulations.

Phase velocities of Rayleigh and Love waves in central and northern Europe from automated, broad-band, interstation measurements

The increasingly dense coverage of Europe with broad-band seismic stations makes it possible to image its lithospheric structure in great detail, provided that structural information can be extracted effectively from the very large volumes of data. We develop an automated technique for the measurement of interstation phase velocities of (earthquake-excited) fundamental-mode surface waves in very broad period ranges. We then apply the technique to all available broad-band data from permanent and temporary networks across Europe. In a new implementation of the classical two-station method, Rayleigh and Love dispersion curves are determined by cross-correlation of seismograms from a pair of stations. An elaborate filtering and windowing scheme is employed to enhance the target signal and makes possible a significantly broader frequency band of the measurements, compared to previous implementations of the method. The selection of acceptable phase-velocity measurements for each event is performed in the frequency domain, based on a number of fine-tuned quality criteria including a smoothness requirement. Between 5 and 3000 single-event dispersion measurements are averaged per interstation path in order to obtain robust, broad-band dispersion curves with error estimates. In total, around 63,000 Rayleigh- and 27,500 Love-wave dispersion curves between 10 and 350 s have been determined, with standard deviations lower than 2 per cent and standard errors lower than 0.5 per cent. Comparisons of phase-velocity measurements using events at opposite backazimuths and the examination of the variance of the phase-velocity curves are parts of the quality control. With the automated procedure, large data sets can be consistently and repeatedly measured using varying selection parameters. Comparison of average interstation dispersion curves obtained with different degrees of smoothness shows that rough perturbations do not systematically bias the average dispersion measurement. They can, therefore, be treated as random but they do need to be removed in order to reduce random errors of the measurements. Using our large new data set, we construct phase-velocity maps for central and northern Europe. According to checkerboard tests, the lateral resolution in central Europe is ≤150 km. Comparison of regional surface-wave tomography with independent data on sediment thickness in North-German Basin and Polish Trough confirms the high-resolution potential of our phase-velocity measurements. At longer periods, the structure of the lithosphere and asthenosphere around the Trans-European Suture Zone (TESZ) is seen clearly. The region of the Tornquist-Teisseyre-Zone in the southeast is associated with a stronger lateral contrast in lithospheric thickness, across the TESZ compared to the region across the Sorgenfrei-Tornquist-Zone in the northwest.

Sparse high-dimensional varying coefficient model: Nonasymptotic minimax study

The objective of the present paper is to develop a minimax theory for the varying coefficient model in a nonasymptotic setting. We consider a high-dimensional sparse varying coefficient model where only few of the covariates are present and only some of those covariates are time dependent. Our analysis allows the time-dependent covariates to have different degrees of smoothness and to be spatially inhomogeneous. We develop the minimax lower bounds for the quadratic risk and construct an adaptive estimator which attains those lower bounds within a constant (if all time-dependent covariates are spatially homogeneous) or logarithmic factor of the number of observations.

High-speed video capture by a single flutter shutter camera using three-dimensional hyperbolic wavelets

Based on the consideration of easy achievement in modern sensors, this paper further exploits the possibility of the recovery of high-speed video (HSV) by a single flutter shutter camera. Taking into account different degrees of smoothness along the spatial and temporal dimensions of HSV, this paper proposes to use a three-dimensional hyperbolic wavelet basis based on Kronecker product to jointly model the spatial and temporal redundancy of HSV. Besides, we incorporate the total variation of temporal correlations in HSV as a prior knowledge to further enhance our reconstruction quality. We recover the underlying HSV frames from the observed low-speed coded video by solving a convex minimization problem. The experimental results on simulated and real-world videos both demonstrate the validity of the proposed method.

Optimal randomized changing dimension algorithms for infinite-dimensional integration on function spaces with ANOVA-type decomposition

We study the numerical integration problem for functions with infinitely many variables. The function spaces of integrands we consider are weighted reproducing kernel Hilbert spaces with norms related to the ANOVA decomposition of the integrands. The weights model the relative importance of different groups of variables. We investigate randomized quadrature algorithms and measure their quality by estimating the randomized worst-case integration error.In particular, we provide lower error bounds for a very general class of randomized algorithms that includes non-linear and adaptive algorithms. Furthermore, we propose new randomized changing dimension algorithms (also called multivariate decomposition methods) and present favorable upper error bounds. For product weights and finite-intersection weights our lower and upper error bounds match and show that our changing dimension algorithms are optimal in the sense that they achieve convergence rates arbitrarily close to the best possible convergence rate. As more specific examples, we discuss unanchored Sobolev spaces of different degrees of smoothness and randomized changing dimension algorithms that use as building blocks interlaced scrambled polynomial lattice rules.

Nonlinear Varying Coefficient Models with Applications to Studying Photosynthesis.

Motivated by a study on factors affecting the level of photosynthetic activity in a natural ecosystem, we propose nonlinear varying coefficient models, in which the relationship between the predictors and the response variable is allowed to be nonlinear. One-step local linear estimators are developed for the nonlinear varying coefficient models and their asymptotic normality is established leading to point-wise asymptotic confidence bands for the coefficient functions. Two-step local linear estimators are also proposed for cases where the varying coefficient functions admit different degrees of smoothness; bootstrap confidence intervals are utilized for inference based on the two-step estimators. We further propose a generalized F test to study whether the coefficient functions vary over a covariate. We illustrate the proposed methodology via an application to an ecology data set and study the finite sample performance by Monte Carlo simulation studies.

2010 Rietz lecture: When does the screening effect hold?

When using optimal linear prediction to interpolate point observations of a mean square continuous stationary spatial process, one often finds that the interpolant mostly depends on those observations located nearest to the predictand. This phenomenon is called the screening effect. However, there are situations in which a screening effect does not hold in a reasonable asymptotic sense, and theoretical support for the screening effect is limited to some rather specialized settings for the observation locations. This paper explores conditions on the observation locations and the process model under which an asymptotic screening effect holds. A series of examples shows the difficulty in formulating a general result, especially for processes with different degrees of smoothness in different directions, which can naturally occur for spatial-temporal processes. These examples lead to a general conjecture and two special cases of this conjecture are proven. The key condition on the process is that its spectral density should change slowly at high frequencies. Models not satisfying this condition of slow high-frequency change should be used with caution.