WAVELET ESTIMATION FOR JUMPS IN A HETEROSCEDASTIC REGRESSION MODEL

11砒roductlonAnalysis ofjumps Is very important Inpractlce．Thejumps often predicts that the in－vestlgated objects are subject to sudden山auges In山aractenstlcs．刊r exaxnple,the jumps ofn 6Xchs,lxge fat6 ill illAnC6 OftCh ShOW th6 lllfiU6DC6 of th6 11POTts;llt 6y6llts h th6 WOTld Oil6nance markt;thejumps ofa seismic signal In oil exploration m叫 imply that there eistsbroken stratum In the expfored strata．It is hot 6My to d6t6Ct th6 JllthPS SlllC6 th6 llld6Ylying Signal Is Oft6l…     

Testing heteroscedasticity by wavelets in a nonparametric regression model

In the nonparametric regression models, a homoscedastic structure is usually assumed. However, the homoscedasticity cannot be guaranteed a priori. Hence, testing the heteroscedasticity is needed. In this paper we propose a consistent nonparametric test for heteroscedasticity, based on wavelets. The empirical wavelet coefficients of the conditional variance in a regression model are defined first. Then they are shown to be asymptotically normal, based on which a test statistic for the heteroscedasticity is constructed by using Fan's wavelet thresholding idea. Simulations show that our test is superior to the traditional nonparametric test.     

Signal analysis based on complex wavelet signs

We propose a signal analysis tool based on the sign (or the phase) of complex wavelet coefficients, which we call a signature. The signature is defined as the fine-scale limit of the signs of a signal's complex wavelet coefficients. We show that the signature equals zero at sufficiently regular points of a signal whereas at salient features, such as jumps or cusps, it is non-zero. At such feature points, the orientation of the signature in the complex plane can be interpreted as an indicator of local symmetry and antisymmetry. We establish that the signature rotates in the complex plane under fractional Hilbert transforms. We show that certain random signals, such as white Gaussian noise and Brownian motions, have a vanishing signature. We derive an appropriate discretization and show the applicability to signal analysis.     

A universal approach to estimate the conditional variance in semimartingale limit theorems

The typical central limit theorems in high-frequency asymptotics for semimartingales are results on stable convergence to a mixed normal limit with an unknown conditional variance. Estimating this conditional variance usually is a hard task, in particular when the underlying process contains jumps. For this reason, several authors have recently discussed methods to automatically estimate the conditional variance, i.e. they build a consistent estimator from the original statistics, but computed at different time scales. Their methods work in several situations, but are essentially restricted to the case of continuous paths always. The aim of this work is to present a new method to consistently estimate the conditional variance which works regardless of whether the underlying process is continuous or has jumps. We will discuss the case of power variations in detail and give insight to the heuristics behind the approach.

     

Generalizing the ENO-DB2<Emphasis Type="Italic">p</Emphasis> transform using the inverse wavelet transform

The essentially non-oscillatory (ENO)-wavelet transform developed by Chan and Zhou (SIAM J. Numer. Anal. 40(4), 1369–1404, 2002) is based on a combination of the Daubechies-2p wavelet transform and the ENO technique. It uses extrapolation methods to compute the scaling coefficients without differencing function values across jumps and obtains a multiresolution framework (essentially) free of edge artifacts. In this work, we present a different way to compute the ENO-DB2p wavelet transform of Chan and Zhou which allows us to simplify the process and easily generalize it to other families of orthonormal wavelets.     

Approximating the jump discontinuities of a function by its Fourier-Jacobi coefficients

In the present paper we generalize Eckhoff's method, i.e., the method for approximating the locations of discontinuities and the associated jumps of a piecewise smooth function by means of its Fourier-Chebyshev coefficients.

A new method enables us to approximate the locations of discontinuities and the associated jumps of a discontinuous function, which belongs to a restricted class of the piecewise smooth functions, by means of its Fourier-Jacobi coefficients for arbitrary indices. Approximations to the locations of discontinuities and the associated jumps are found as solutions of algebraic equations. It is shown as well that the locations of discontinuities and the associated jumps are recovered exactly for piecewise constant functions with a finite number of discontinuities.

In addition, we study the accuracy of the approximations and present some numerical examples.

     

Shrinkage estimator in normal mean vector estimation based on conditional maximum likelihood estimators

Estimation of normal mean vector has broad applications such as small area estimation, estimation of nonparametric functions and estimation of wavelet coefficients. In this paper, we propose a new shrinkage estimator based on conditional maximum likelihood estimator incorporating with Stein’s risk unbiased estimator (SURE) when data have the normality. We present some theoretical work and provide numerical studies to compare with some existing methods.     

JUMP DETECTION BY WAVELET IN NONLINEAR AUTOREGRESSIVE MODELS

1991MRSubjectClassification62G05,62G201IntroductionDtttectiollofthe.iulnppointshasrecentlyfoundinCleasillginterests.Sincejllliippoillts(\andftstfriheson-iesuddenchallgephenorxlenonena,theyal'every11seflllillrllodellillgpracticalprobl(!lusarisinginfieldssuchaseconomics,signalanalysis,illlageprocessingandphonetici'lentification.TheeallyworkondetectiollofthejumpsisShi..[1]andSpeckman[2].Yin[']consideredthe1llodely(t)=s(t) e(f),05t51,(1.1)wheree(t)isaGaussianwhitenoisewithe(0)=0ands(f)isadeter…     

ASYMPTOTIC BEHAVIOR OF THE ECKHOFF APPROXIMATION IN BIVARIATE CASE

The paper considers the Krylov-Lanczos and the Eckhoff approximations for recovering a bivariate function using limited number of its Fourier coefficients. These approximations are based on certain corrections associated with jumps in the partial derivatives of the approximated function. Approximation of the exact jumps is accomplished by solution of systems of linear equations along the idea of Eckhoff. Asymptotic behaviors of the approximate jumps and the Eckhoff approximation are studied. Exact constants of the asymptotic errors are computed. Numerical experiments validate theoretical investigations.     

Large scale behavior of wavelet coefficients of non-linear subordinated processes with long memory

We study the asymptotic behavior of wavelet coefficients of random processes with long memory. These processes may be stationary or not and are obtained as the output of non-linear filter with Gaussian input. The wavelet coefficients that appear in the limit are random, typically non-Gaussian and belong to a Wiener chaos. They can be interpreted as wavelet coefficients of a generalized self-similar process.     

Towards artifact-free characterization of surface topography using complex wavelets and total variation minimization

A hybrid method based on dual-tree complex wavelet transform and total variation minimization is proposed for erasure of undesirable artifacts that arise in the existing wavelet-based methods in surface metrology. The complex wavelet transform provides approximate shift invariance and good directional selectivity, and attempts to solve the weakness of real discrete wavelet methods. Reconstruct the complex wavelet coefficients using a total variation minimization principle to eliminate the wavelet-shape artifacts and the pseudo-Gibbs artifacts near the discontinuities, which are caused when thresholding small wavelet coefficients. By replacing these thresholded complex wavelet coefficients by optimal values that minimize the total variation, the method performs a close artifact-free surface characterization. Numerical experiments using a series of typical engineering and bioengineering surfaces demonstrate the remarkable potential of the methodology.     

Additive Schwarz Method for Mortar Discretization of Elliptic Problems with <Emphasis Type="Italic">P</Emphasis><Subscript>1</Subscript> Nonconforming Finite Elements

An additive Schwarz preconditioner for nonconforming mortar finite element discretization of a second order elliptic problem in two dimensions with arbitrary large jumps of the discontinuous coefficients in subdomains is described. An almost optimal estimate of the condition number of the preconditioned problem is proved. The number of preconditioned conjugate gradient iterations is independent of jumps of the coefficients and is proportional to (1+log(H/h)), where H,h are mesh sizes. AMS subject classification (2000) 65N55, 65N30, 65N22     

BOOTSTRAP WAVELET IN THE NONPARAMETRIC REGRESSION MODEL WITH WEAKLY DEPENDENT PROCESSES

This paper introduces a method of bootstrap wavelet estimation in a nonparametric regression model with weakly dependent processes for both fixed and random designs. The asymptotic bounds for the bias and variance of the bootstrap wavelet estimators are given in the fixed design model. The conditional normality for a modified version of the bootstrap wavelet estimators is obtained in the fixed model. The consistency for the bootstrap wavelet estimator is also proved in the random design model. These results show that the bootstrap wavelet method is valid for the model with weakly dependent processes.     

适应于图像压缩的11/9小波基构造

根据正交多分辨分析理论,利用求解低通和高通滤波的系数,可构造出多种正交小波.但正交小波中只有Haar小波满足对称性,这不适合在图像处理方面的应用.在提升格式的小波变换出现之前,小波分解通过Mallat算法来完成,而提升格式的小波有显著的优点,运算量少,不同小波运算量减少程度不一样,一般减少在25%到50%之间.文章根据双正交对称紧支集小波的消失矩、对称性、短支撑等一系列条件和其他构造原理,构造出一个适应图像压缩的11/9双正交提升小波,并满足Cohen-Daubechies准则.同时,为了便于小波变换的硬件实现,最佳的状态是,分解和重构滤波系数为二进制分数,且根据不同参数取值,让子带编码增益G_(SBC)达到最大.     

Exponential stability of Hopfield neural networks with impulses

For a Hopfield neural network with periodic coefficients, a new criterion is proposed to obtain the existence of a periodic solution and its exponential stability. Our assumptions are in the form of inequalities involving integral averages and the assigned jumps.     

The wavelet detection of the jump and cusp points of a regression function

1. IntroductionMuch effort has been taken to detect the change points of a noise contaminated signal. Detection of change points is very useful in dealing with practical problems such assignal analysis, image processing and phonetic identification. For example, in dealing withelect ro encep halogr am signal ? do ct ors of t en need t o find re al sharp cusp s which exhibi t t heaccelerations and decelerations in the beating of hearts. The early work on detection ofthe change points of a regres…     

Wavelet Approximation of Periodic Functions

We investigate expansions of periodic functions with respect to wavelet bases. Direct and inverse theorems for wavelet approximation in C and L_p norms are proved. For the functions possessing local regularity we study the rate of pointwise convergence of wavelet Fourier series. We also define and investigate the “discreet wavelet Fourier transform” (DWFT) for periodic wavelets generated by a compactly supported scaling function. The DWFT has one important advantage for numerical problems compared with the corresponding wavelet Fourier coefficients: while fast computational algorithms for wavelet Fourier coefficients are recursive, DWFTs can be computed by explicit formulas without any recursion and the computation is fast enough.     

Stochastic differential equations with diffusion and jumps modeling currency markets

An efficient currency market with zero transaction costs is considered. The dynamics of the exchange rate in this market is described by stochastic differential equations (SDEs) with diffusion and jumps; the latter are assumed to be described by a Lévy process. Adjusting theoretical arbitrage-free option prices computed within these models to market option prices requires properly choosing the coefficients in the SDEs. For this purpose, an expression for local volatility in a diffusion model is found and a relation between local and implied volatilities is determined. For a market model with diffusion and jumps, expressions for the local volatility and the local rate function are given. Moreover, in Merton’s model, where the jump component is a compound Poisson process with normal jumps, a relation between the local and the implied volatilities is determined.     

Adaptive quantile regression with precise risk bounds

An adaptive local smoothing method for nonparametric conditional quantile regression models is considered in this paper. Theoretical properties of the procedure are examined. The proposed method is fully adaptive in the sense that no prior information about the structure of the model is assumed. The fully adaptive feature not only allows varying bandwidths to accommodate jumps or instantaneous slope changes, but also allows the algorithm to be spatially adaptive. Under general conditions, precise risk bounds for homogeneous and heterogeneous cases of the underlying conditional quantile curves are established. An automatic selection algorithm for locally adaptive bandwidths is also given, which is applicable to higher dimensional cases. Simulation studies and data analysis confirm that the proposed methodology works well.     

Asymptotic behavior of the threshold minimizing the average probability of error in calculation of wavelet coefficients

The problem of estimating the signal function from noisy observations by thresholding the coefficients of its wavelet decomposition is considered. The asymptotic orders of the threshold and risk are calculated by minimizing the average probability of error in calculating the wavelet coefficients.