Brani Vidakovic

Nonlinear Wavelet Shrinkage with Bayes Rules and Bayes Factors

Abstract Wavelet shrinkage, the method proposed by the seminal work of Donoho and Johnstone is a disarmingly simple and efficient way of denoising data. Shrinking wavelet coefficients was proposed from several optimality criteria. In this article a wavelet shrinkage by coherent Bayesian inference in the wavelet domain is proposed. The methods are tested on standard Donoho-Johnstone test functions.

The Effectiveness of Sealants in Managing Caries Lesions

A barrier to providing sealants is concern about inadvertently sealing over caries. This meta-analysis examined the effectiveness of sealants in preventing caries progression. We searched electronic databases for comparative studies examining caries progression in sealed permanent teeth. We used a random-effects model to estimate percentage reduction in the probability of caries progression in sealed vs. unsealed carious teeth. Six studies, including 4 randomized-controlled trials (RCT) judged to be of fair quality, were included in the analysis (384 persons, 840 teeth, and 1090 surfaces). The median annual percentage of non-cavitated lesions progressing was 2.6% for sealed and 12.6% for unsealed carious teeth. The summary prevented fraction for RCT was 71.3% (95%CI: 52.8%–82.5, no heterogeneity) up to 5 years after placement. Despite variation among studies in design and conduct, sensitivity analysis found the effect to be consistent in size and direction. Sealing non-cavitated caries in permanent teeth is effective in reducing caries progression.

Multiscale analysis of vegetation surface fluxes: from seconds to years

Abstract The variability in land surface heat (H), water vapor (LE), and CO2 (or net ecosystem exchange, NEE) fluxes was investigated at scales ranging from fractions of seconds to years using eddy-covariance flux measurements above a pine forest. Because these fluxes significantly vary at all these time scales and because large gaps in the record are unavoidable in such experiments, standard Fourier expansion methods for computing the spectral and cospectral statistical properties were not possible. Instead, orthonormal wavelet transformations ( OWT ) are proposed and used. The OWT are ideal at resolving process variability with respect to both scale and time and are able to isolate and remove the effects of missing data (or gaps) from spectral and cospectral calculations. Using the OWT spectra, we demonstrated unique aspects in three appropriate ranges of time scales: turbulent time scales (fractions of seconds to minutes), meteorological time scales (hour to weeks), and seasonal to interannual time scales corresponding to climate and vegetation dynamics. We have shown that: (1) existing turbulence theories describe the short time scales well, (2) coupled physiological and transport models (e.g. CANVEG) reproduce the wavelet spectral characteristics of all three land surface fluxes for meteorological time scales, and (3) seasonal dynamics in vegetation physiology and structure inject strong correlations between land surface fluxes and forcing variables at monthly to seasonal time scales. The broad implications of this study center on the possibility of developing low-dimensional models of land surface water, energy, and carbon exchange. If the bulk of the flux variability is dominated by a narrow band or bands of modes, and these modes “resonate” with key state and forcing variables, then low-dimensional models may relate these forcing and state variables to NEE and LE.

Nonparametric Statistics with Applications to Science and Engineering: Kvam/Nonparametric Statistics

Preface. 1. Introduction. 2. Probability Basics. 3. Statistics Basics. 4. Bayesian Statistics. 5. Order Statistics. 6. Goodness of Fit. 7. Rank Tests. 8. Designed Experiments. 9. Categorical Data. 10. Estimating Distribution Functions. 11. Density Estimation. 12. Beyond Linear Regression. 13. Curve Fitting Techniques. 14. Wavelets. 15. Bootstrap. 16. EM Algorithm. 17. Statistical Learning. 18. Nonparametric Bayes. A. MATLAB. B. WinBUGS. MATLAB Index. Author Index. Subject Index.

Bayesian inference in wavelet-based models

I Introduction.- 1 An Introduction to Wavelets.- 2 Spectral View of Wavelets and Nonlinear Regression.- II Prior Models - Independent Case.- 3 Bayesian Approach to Wavelet Decomposition and Shrinkage.- 4 Some Observations on the Iractability of Certain Multi-Scale Models..- 5 Bayesian Analysis of Change-Point Models.- 6 Prior Elicitation in the Wavelet Domain.- 7 Wavelet Nonparametric Regression Using Basis Averaging.- III Decision Theoretic Wavelet Shrinkage.- 8 An Overview of Wavelet Regularization.- 9 Minimax Restoration and Deconvolution.- 10 Robust Bayesian and Bayesian Decision Theoretic Wavelet Shrinkage.- 11 Best Basis Representations with Prior Statistical Models.- IV Prior Models - Dependent Case.- 12 Modeling Dependence in the Wavelet Domain.- 13 MCMC Methods in Wavelet Shrinkage.- V Spatial Models.- 14 Empirical Bayesian Spatial Prediction Using Wavelets.- 15 Geometrical Priors for Noisefree Wavelet Coefficients in Image Denoising.- 16 Multiscale Hidden Markov Models for Bayesian Image Analysis.- 17 Wavelets for Object Representation and Recognition in Computer Vision.- 18 Bayesian Denoising of Visual Images in the Wavelet Domain.- VI Empirical Bayes.- 19 Empirical Bayes Estimation in Wavelet Nonparametric Regression.- 20 Nonparametric Empirical Bayes Estimation via Wavelets.- VII Case Studies.- 21 Multiresolution Wavelet Analyses in Hierarchical Bayesian Turbulence Models.- 22 Low Dimensional Turbulent Transport Mechanics Near the Forest-Atmosphere Interface.- 23 Latent Structure Analyses of Turbulence Data Using Wavelets and Time Series Decompositions.

Active Turbulence and Scalar Transport near the Forest–Atmosphere Interface

Turbulent velocity, temperature, water vapor concentration, and other scalars were measured at the canopy‐ atmosphere interface of a 13‐14-m-tall uniform pine forest and a 33-m-tall nonuniform hardwood forest. These measurements were used to investigate whether the mixing layer (ML) analogy of Raupach et al. predicts eddy sizes and flow characteristics responsible for much of the turbulent stresses and vertical scalar fluxes. For this purpose, wavelet spectra and cospectra were derived and analyzed. It was found that the ML analogy predicts well vertical velocity variances and integral timescales. However, at low wavenumbers, inactive eddy motion signatures were present in horizontal velocity wavelet spectra, suggesting that ML may not be suitable for scaling horizontal velocity perturbations. Momentum and scalar wavelet cospectra of turbulent stresses and scalar fluxes demonstrated that active eddy motion, which was shown by Raupach et al. to be the main energy contributor to vertical velocity (w) spectral energy (Ew), is also the main scalar flux‐transporting eddy motion. Predictions using ML of the peak Ew frequency are in excellent agreement with measured wavelet cospectral peaks of vertical fluxes ( Kh 5 1.5, where K is wavenumber and h is canopy height). Using Lorentz wavelet thresholding of vertical velocity time series, wavelet coefficients associated with active turbulence were identified. It was demonstrated that detection frequency of organized structures, as predicted from Lorentz wavelet filtering, relate to the arrival frequency ^U&/h and integral timescale, where ^U& is the mean horizontal velocity at height z 5 h. The newly proposed wavelet thresholding approach, which relies on a ‘‘global’’ wavelet threshold formulation for the energy in w, provides simultaneous energy‐covariance-preserving characterization of ‘‘active’’ turbulence at the canopy‐atmosphere interface.

The partitioning of attached and detached eddy motion in the atmospheric surface layer using Lorentz wavelet filtering

Townsends attached eddy hypothesis states that the turbulent structure in the constant stress layer can be decomposed into attached and detached eddy motion. This paper proposes and tests a methodology for separating the attached and detached eddy motion from time series measurements of velocity and temperature. The proposed methodology is based on the time-frequency localization and filtering capabilities of the orthonormal wavelet transforms. Using a relative entropy statistical measure, the optimal wavelet basis is identified first. The turbulence time series measurements are then transformed into the wavelet domain where the contribution of specific events in the time-frequency domain is identified. The filtering scheme utilizes a recently constructed Lorentz thresholding methodology that successfully eliminates all wavelet coefficients associated with the detached eddy motion. While this filtering scheme lacks the compression efficiency of the classical Donoho and Johnstones universal thresholding model, it conserves the higher-order statistics and important turbulence interactions related to the Reynolds stresses. Following the filtering scheme, the attached eddy motion time series is re-constructed by an inverse wavelet transform of the non-zero wavelet coefficients. The proposed partitioning methodology for attached and detached eddy motion is tested using 56 Hz triaxial sonic anemometer velocity and temperature measurements above a uniform dry lake bed in Owens valley, California, for a wide range of atmospheric stability conditions. Validation that the wavelet filtered time series represents the attached eddy motion is also discussed in the context of conservation of turbulence energy and surface fluxes.

Whole-body immunoPET reveals active SIV dynamics in viremic and antiretroviral therapy–treated macaques

The detection of viral dynamics and localization in the context of controlled HIV infection remains a challenge and is limited to blood and biopsies. We developed a method to capture total-body simian immunodeficiency virus (SIV) replication using immunoPET (antibody-targeted positron emission tomography). The administration of a poly(ethylene glycol)-modified, 64Cu-labeled SIV Gp120–specific antibody led to readily detectable signals in the gastrointestinal and respiratory tract, lymphoid tissues and reproductive organs of viremic monkeys. Viral signals were reduced in aviremic antiretroviral-treated monkeys but detectable in colon, select lymph nodes, small bowel, nasal turbinates, the genital tract and lung. In elite controllers, virus was detected primarily in foci in the small bowel, select lymphoid areas and the male reproductive tract, as confirmed by quantitative reverse-transcription PCR (qRT-PCR) and immunohistochemistry. This real-time, in vivo viral imaging method has broad applications to the study of immunodeficiency virus pathogenesis, drug and vaccine development, and the potential for clinical translation.

Hotel general manager career paths in the United States

Abstract The career paths of 114 general managers (GM) of U.S. mid-range, up-scale and luxury hotels were studied. Nearly 80% held the Assistant GM position before becoming GMs. GMs followed narrow career paths prior to becoming Assistant GMs, spending 87.5% of their time in only one department. A hotels two major operational departments, F&B and Rooms, accounted for three-quarters of the GM career paths, with F&B representing 45%. Career paths remained stable for the three types of hotels. Career paths of younger GMs have shifted decidedly towards these two operational departments: 85.4% of the younger half followed career paths in either F&B or Rooms compared to 65% of the older GMs.

Skin temperature perturbations induced by surface layer turbulence above a grass surface

High-frequency (5 Hz) atmospheric surface layer (ASL) turbulent velocity (u′) and infrared skin temperature perturbations (T′s) were measured above a grass-covered forest clearing and analyzed for cloud free conditions. These measurements were used to investigate mechanisms responsible for the production of large short-lived T′s perturbations caused by rapid excursions in u′. To quantify the effects of u′ on rapid surface cooling, wavelet spectra of u′ and T′s and cospectra of u′T′s were computed. The u′ wavelet power spectra were then analyzed using Townsends [1961, 1976] hypothesis. Townsends hypothesis states that ASL eddy motion can be decomposed into an active component, which is a function of the ground shear stress (u*) and height (z) above the zero plane displacement, and an inactive component, which is produced in the atmospheric boundary layer (ABL) outer region. A −1 power law in the u′ power spectrum was used as a signature for inactive eddy motion. Therefore the −1 power law was used to identify wavenumber ranges (about 1.5 decades) associated with inactive eddy motion. The measured T′s wavelet spectra and u′T′s cospectra identified with this wavenumber range demonstrate that much of the T′s energy and 〈u′T′s〉 are due to inactive eddy motion, where the angle brackets indicate time averaging. Hence, in contrast to the laboratory experiments of Owen and Thomson [1963], it is argued that skin temperature perturbations at the canopy-atmosphere interface of a grass-covered surface (small thermal inertia) are strongly dependent on the inactive eddy motion produced in the outer layer of the ABL.