Search

Buffelgrass [Pennisetum ciliare (L.) Link] is an invasive C4 perennial bunchgrass that is a threat to biodiversity in aridlands in the Americas and Australia. Topography influences P. ciliare occurrence at large spatial scales, but further investigation into the relationship between local-scale topography and P. ciliare growth and reproduction would be beneficial. Further, density-dependent effects on P. ciliare growth and reproduction have been demonstrated in greenhouse experiments, but the extent to which density dependence influences P. ciliare in natural populations warrants further investigation. Here we present a study on the relationships between local-scale topography (aspect and slope gradient) and vegetation characteristics (shrub cover, P. ciliare cover, and P. ciliare density) and their interactions on individual P. ciliare plant size and reproduction. We measured slope gradient, aspect, shrub cover, P. ciliare cover, P. ciliare density, and the total number of live culms and reproductive culms of 10 P. ciliare plants in 33 4 by 4 m plots located in 11 transects at the Desert Laboratory at Tumamoc Hill, Tucson, AZ, USA. We modeled the relationships at the local scale of (1) P. ciliare cover and density with aspect and slope gradient and (2) P. ciliare size and reproduction with abiotic (slope gradient and aspect) and biotic (P. ciliare cover and density and native shrub and cacti cover) characteristics. Aspect and slope gradient were poor predictors of P. ciliare cover and density in already invaded sites at the scale of our plots. However, aspect had a significant relationship with P. ciliare plant size and reproduction. Pennisetum ciliare plants on south-facing aspects were larger and produced more reproductive culms than plants on other aspects. Further, we found no relationship between P. ciliare density and P. ciliare plant size and reproduction. Shrub cover was positively correlated with P. ciliare reproduction. South-facing aspects are likely most vulnerable to fast spread and infilling by new P. ciliare introductions.

We prove a universality theorem for learning with random features. Our result shows that, in terms of training and generalization errors, a random feature model with a nonlinear activation function is asymptotically equivalent to a surrogate linear Gaussian model with a matching covariance matrix. This settles a so-called Gaussian equivalence conjecture based on which several recent papers develop their results. Our method for proving the universality theorem builds on the classical Lindeberg approach. Major ingredients of the proof include a leave-one-out analysis for the optimization problem associated with the training process and a central limit theorem, obtained via Stein's method, for weakly correlated random variables.

The invasion of nonnative grasses threatens biodiversity and ecosystem function globally through competition with native plant species and increases to wildfire frequency and intensity. Management actions to reduce buffelgrass [Pennisetum ciliare (L.) Link], an invasive warm-season perennial bunchgrass, are widely implemented, with chemical and mechanical treatments extending over two decades within Saguaro National Park in the Sonoran Desert of North America. We assessed how the effectiveness of treatments to reduce P. ciliare cover spanning from 2011 to 2020 were influenced by stage of invasion, treatment type and intensity, and environmental conditions. An increase in treatment effectiveness was largely explained by high initial cover of P. ciliare, an indicator of a late invasion stage and associated with high treatment intensity. Treatments had potential to be effective in patches as small as 0.3-m2P. ciliare canopy per 400-m−2 area (P. ciliare were associated with shorter average years of treatment interruptions, or gaps, and to a lesser degree, total years of treatment. In many cases, P. ciliare was reduced with as little as 2 yr of treatment, but more than 3 average years of treatment gap could result in reduced treatment effectiveness. There was generally higher treatment effectiveness on shallow slopes, north- and east-facing aspects, and on higher elevations within one district of the park. Our findings highlight that resource-intensive treatments in all but the smallest patches of P. ciliare have largely been effective. Further opportunities for improvement include more frequent surveillance, limiting treatment gaps to ≤3 yr in areas of low P. ciliare cover, and comparison of treated with untreated areas.

Substantial gaps exist between weed management researchers and practitioners with respect to prompt exchange of knowledge between the two groups, hindering the implementation of effective management to solve weed problems. We conducted a survey between 2016 and 2018 among weed management practitioners (n = 259) across diverse ecoregions on California rangelands and collected essential information from practitioners for bridging the research-implementation gap. The information included management costs, high-priority weeds, and spatial scales and temporal changes in weed management. The management cost had a mean of $5.12 ha−1yr−1 with a large variance implying the uncertainty of this information. The percentage of annual budget dedicated to weed management explained about 30% of the variation in this cost. Moreover, this annual per-unit area cost decreased with increasing management area. The average size of rangeland managed by survey respondents was 1 256 ha. The top three high-priority weeds statewide were yellow starthistle (Centaurea solstitialis), medusahead (Taeniatherum caput-medusae), and all thistles combined. Medusahead and some thistle species remained on the top list in each ecoregion. Respondents overwhelmingly (80.9%) noted changes in weed problems in the past 5−10 yr, specifically citing greater weed pressure and changes in weed species. A significantly higher proportion of respondents from agencies and private businesses reported changes in weed problems than those from universities and nongovernmental organizations (NGOs), further underscoring the gap between practitioners and researchers. A majority of the respondents (75.3%) indicated the record-setting California drought had a negative effect on weed management, reducing treatment efficacy and favoring weeds over desirable species. Overall, our findings illustrate increasing challenges in weed management on California rangeland. These challenges call for adaptive management-research programs to increase cost-effectiveness of weed management and to swiftly and effectively respond to dynamic weed problems on large spatial and long temporal scales.

Thermal barrier coating is a high-temperature protective technology widely used in industrial gas turbines. However, the failure of coating peeling because of the generation of thermally grown oxide (TGO) at the interface during service hinders its further application. In this study, Raman spectroscopy and wedge indentation are used to determine the TGO residual stress and the interface energy release rate, respectively. The effect of TGO on the interfacial fracture toughness during the growth process was discussed. Raman spectroscopy test results show that the residual stress of TGO is about 0.5 GPa. Wedge indentation test results illustrate that high-temperature heat treatment could accelerate the interface degradation of thermal barrier coatings. Stress analysis and test research demonstrate that the microcracks induced by compressive stress of TGO will propagate with increasing heating time, ending with failure of barrier coatings.

We analyze the dynamics of a random sequential message passing algorithm for approximate inference with large Gaussian latent variable models in a student–teacher scenario. To model nontrivial dependencies between the latent variables, we assume random covariance matrices drawn from rotation invariant ensembles. Moreover, we consider a model mismatching setting, where the teacher model and the one used by the student may be different. By means of dynamical functional approach, we obtain exact dynamical mean-field equations characterizing the dynamics of the inference algorithm. We also derive a range of model parameters for which the sequential algorithm does not converge. The boundary of this parameter range coincides with the de Almeida Thouless (AT) stability condition of the replica-symmetric ansatz for the static probabilistic model.

Approximate Message Passing (AMP) is a class of iterative algorithms that have found applications in many problems in high-dimensional statistics and machine learning. In its general form, AMP can be formulated as an iterative procedure driven by a matrix $\mathbf{M}$. Theoretical analyses of AMP typically assume strong distributional properties on $\mathbf{M}$ such as $\mathbf{M}$ has i.i.d. sub-Gaussian entries or is drawn from a rotational invariant ensemble. However, numerical experiments suggest that the behavior of AMP is universal, as long as the eigenvectors of $\mathbf{M}$ are generic. In this paper, we take the first step in rigorously understanding this universality phenomenon. In particular, we investigate a class of memory-free AMP algorithms (proposed by \c{C}akmak and Opper for mean-field Ising spin glasses), and show that their asymptotic dynamics is universal on a broad class of semi-random matrices. In addition to having the standard rotational invariant ensemble as a special case, the class of semi-random matrices that we define in this work also includes matrices constructed with very limited randomness. One such example is a randomly signed version of the Sine model, introduced by Marinari, Parisi, Potters, and Ritort for spin glasses with fully deterministic couplings.

We investigate random matrices whose entries are obtained by applying a nonlinear kernel function to pairwise inner products between $n$ independent data vectors, drawn uniformly from the unit sphere in $\mathbb{R}^d$. This study is motivated by applications in machine learning and statistics, where these kernel random matrices and their spectral properties play significant roles. We establish the weak limit of the empirical spectral distribution of these matrices in a polynomial scaling regime, where $d, n \to \infty$ such that $n / d^\ell \to \kappa$, for some fixed $\ell \in \mathbb{N}$ and $\kappa \in (0, \infty)$. Our findings generalize an earlier result by Cheng and Singer, who examined the same model in the linear scaling regime (with $\ell = 1$). Our work reveals an equivalence principle: the spectrum of the random kernel matrix is asymptotically equivalent to that of a simpler matrix model, constructed as a linear combination of a (shifted) Wishart matrix and an independent matrix sampled from the Gaussian orthogonal ensemble. The aspect ratio of the Wishart matrix and the coefficients of the linear combination are determined by $\ell$ and the expansion of the kernel function in the orthogonal Hermite polynomial basis. Consequently, the limiting spectrum of the random kernel matrix can be characterized as the free additive convolution between a Marchenko-Pastur law and a semicircle law. We also extend our results to cases with data vectors sampled from isotropic Gaussian distributions instead of spherical distributions.

As modern machine learning models continue to advance the computational frontier, it has become increasingly important to develop precise estimates for expected performance improvements under different model and data scaling regimes. Currently, theoretical understanding of the learning curves that characterize how the prediction error depends on the number of samples is restricted to either large-sample asymptotics ($m\to\infty$) or, for certain simple data distributions, to the high-dimensional asymptotics in which the number of samples scales linearly with the dimension ($m\propto d$). There is a wide gulf between these two regimes, including all higher-order scaling relations $m\propto d^r$, which are the subject of the present paper. We focus on the problem of kernel ridge regression for dot-product kernels and present precise formulas for the mean of the test error, bias, and variance, for data drawn uniformly from the sphere with isotropic random labels in the $r$th-order asymptotic scaling regime $m\to\infty$ with $m/d^r$ held constant. We observe a peak in the learning curve whenever $m \approx d^r/r!$ for any integer $r$, leading to multiple sample-wise descent and nontrivial behavior at multiple scales., Comment: 42 pages; 5 + 6 figures

It has been observed that the performances of many high-dimensional estimation problems are universal with respect to underlying sensing (or design) matrices. Specifically, matrices with markedly different constructions seem to achieve identical performance if they share the same spectral distribution and have ``generic'' singular vectors. We prove this universality phenomenon for the case of convex regularized least squares (RLS) estimators under a linear regression model with additive Gaussian noise. Our main contributions are two-fold: (1) We introduce a notion of universality classes for sensing matrices, defined through a set of deterministic conditions that fix the spectrum of the sensing matrix and precisely capture the previously heuristic notion of generic singular vectors; (2) We show that for all sensing matrices that lie in the same universality class, the dynamics of the proximal gradient descent algorithm for solving the regression problem, as well as the performance of RLS estimators themselves (under additional strong convexity conditions) are asymptotically identical. In addition to including i.i.d. Gaussian and rotational invariant matrices as special cases, our universality class also contains highly structured, strongly correlated, or even (nearly) deterministic matrices. Examples of the latter include randomly signed versions of incoherent tight frames and randomly subsampled Hadamard transforms. As a consequence of this universality principle, the asymptotic performance of regularized linear regression on many structured matrices constructed with limited randomness can be characterized by using the rotationally invariant ensemble as an equivalent yet mathematically more tractable surrogate.

Antibiotics are commonly used to treat infectious diseases. However, massive and inappropriate antibiotics usage cause many problems including the emergence and spread of antibiotic-resistant bacteria. To avoid this issue, in modern countries the interest of using probiotics in feed supplementation to promote health and prevent or treat intestinal infectious diseases in companion animals like dogs has been increasing. We evaluate the probiotic potential of Lactobacilli isolated from healthy dogs faeces. The isolated Lactobacilli were first confirmed by 16SrRNA sequencing, then in vitro tests were conducted to assess survival potential of Lactobacilli under simulated gastrointestinal conditions and adhesion ability to gut epithelia, effects on epithelial barrier function, anti-inflammatory activities, effects on defensin peptides (beta-defensin 3) and inhibitory effects on common pathogens. Lactobacilli showed considerable potential to survive in simulated gastrointestinal environmental conditions, low pH, high bile salt concentrations along with good adhesion properties with MODE-K cells. Pathogenic bacterial growth and their adhesion to MODE-K cells was significantly inhibited by Lactobacilli. Real-time PCR analyses further demonstrated that L. acidophilus strain AR1 and AR3 inhibit Salmonella-induced proinflammatory cytokines (IL-6, IL-8, 1ꞵ) production and reinforce expression of tight junction protein (occludin). None of the strain induce mRNA expression of beta-defensin 3 in MODE-K cells. Based on in vitro results the L. acidophilus strain AR1 has potential to be supplemented in canine feed. However, further in vivo studies investigating health-promoting effects are awaited.

Large-scale control of invasive plants can benefit strongly from reliable assessment of spatial variation in plant invasibility. With this knowledge, limited management resources can be concentrated in areas of high invasion risk. We assessed the influence of spatial environments and proximity to roads on the invasibility of African mustard (Brassica tournefortiiGouan) over the 280,000-ha Barry M. Goldwater Range West in southwestern Arizona, USA. We used presence/absence data ofB. tournefortiiacquired from a vegetation classification project, in which lands were mapped to the level of vegetation subassociations. Logistic regression models suggested that spatial environments represented by the subassociations, not proximity to roads, represented the only factor significantly explainingB. tournefortiipresence. We then used the best model to predictB. tournefortiiinvasibility in each subassociation. This prediction indicates management strategy should differ between the western part and the central to eastern part of the range. The western range is a large spatial continuum with intermediate to high invasion risk, vulnerable to an untethered spread ofB. tournefortii. Controlling efforts should focus on preventing existing local populations from further expansion. The central and eastern ranges are a mosaic varying strongly in invasion risk. Control efforts can take advantage of natural invasion barriers and further reduce connectivity through removal of source populations connected with other high-risk locations via roads and other dispersal corridors. We suggest our approach as one effective way to combine vegetation classification and plant invasion assessment to manage complex landscapes over large ranges, especially when this approach is used through an iterative prediction–validation process to achieve adaptive management of invasive plants.

As enveloped virus, SARS-CoV-2 membrane protein (M) mediates viral release from cellular membranes, but the molecular mechanisms of SARS-CoV-2 virions release remain poorly understood. Here, we performed RNAi screening and identified the E3 ligase RNF5 which mediates ubiquitination of SARS-CoV-2 M at residue K15 to enhance the interaction of viral envelope (E) with M. M-E complex ensures the uniform size of viral particles for viral maturation and mediates viral release. Moreover, overexpression of M induces complete autophagy which is dependent on RNF5-mediated ubiquitin modification. M inhibits the activity of lysosome protease, and uses autolysosomes for virion release. Consequently, all these results demonstrate that RNF5 mediates ubiquitin modification of SARS-CoV-2 M to stabilize the M-E complex and induce autophagy for virion release.

This systematic review provides an overview of the effects of menopausal symptom treatment options on palpitations, defined as feelings of missed or exaggerated heart beats, reported by perimenopausal and postmenopausal women. Guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, searches were conducted in PubMed, CINAHL and PsycINFO to identify articles meeting pre-specified inclusion criteria. Of 670 unique articles identified, 37 were included in the review. Treatments included drug therapies and non-drug therapies. Palpitations were studied as an outcome in 89% of articles and as an adverse effect in 11%. Articles provided mostly level II/III evidence due to their design and/or small sample sizes. Based on available evidence, no therapies can be fully recommended for clinical practice. Only some hormonal agents (e.g. estradiol) can be recommended with caution based on some positive evidence for reducing palpitation prevalence or severity. However, other drug therapies (e.g. moxonidine, atenolol), dietary supplementary treatments (e.g. isoflavones, Rheum rhaponticum, sage), cognitive-behavioral intervention and auricular acupressure cannot be recommended given the existing evidence. Additional well-designed randomized controlled treatment trials focusing on palpitations during the menopause transition as an inclusion criteria and outcome are needed to advance the field.

This paper proposes a new algorithm, named Householder Dice (HD), for simulating dynamics on dense random matrix ensembles with translation-invariant properties. Examples include the Gaussian ensemble, the Haar-distributed random orthogonal ensemble, and their complex-valued counterparts. A "direct" approach to the simulation, where one first generates a dense $n \times n$ matrix from the ensemble, requires at least $\mathcal{O}(n^2)$ resource in space and time. The HD algorithm overcomes this $\mathcal{O}(n^2)$ bottleneck by using the principle of deferred decisions: rather than fixing the entire random matrix in advance, it lets the randomness unfold with the dynamics. At the heart of this matrix-free algorithm is an adaptive and recursive construction of (random) Householder reflectors. These orthogonal transformations exploit the group symmetry of the matrix ensembles, while simultaneously maintaining the statistical correlations induced by the dynamics. The memory and computation costs of the HD algorithm are $\mathcal{O}(nT)$ and $\mathcal{O}(nT^2)$, respectively, with $T$ being the number of iterations. When $T \ll n$, which is nearly always the case in practice, the new algorithm leads to significant reductions in runtime and memory footprint. Numerical results demonstrate the promise of the HD algorithm as a new computational tool in the study of high-dimensional random systems.

We consider a commonly studied supervised classification of a synthetic dataset whose labels are generated by feeding a one-layer neural network with random iid inputs. We study the generalization performances of standard classifiers in the high-dimensional regime where $\alpha=n/d$ is kept finite in the limit of a high dimension $d$ and number of samples $n$. Our contribution is three-fold: First, we prove a formula for the generalization error achieved by $\ell_2$ regularized classifiers that minimize a convex loss. This formula was first obtained by the heuristic replica method of statistical physics. Secondly, focussing on commonly used loss functions and optimizing the $\ell_2$ regularization strength, we observe that while ridge regression performance is poor, logistic and hinge regression are surprisingly able to approach the Bayes-optimal generalization error extremely closely. As $\alpha \to \infty$ they lead to Bayes-optimal rates, a fact that does not follow from predictions of margin-based generalization error bounds. Third, we design an optimal loss and regularizer that provably leads to Bayes-optimal generalization error., Comment: 11 pages + 45 pages Supplementary Material / 5 figures, v2 revised and accepted at NeurIPS

We consider the phase retrieval problem, in which the observer wishes to recover a $n$-dimensional real or complex signal $\mathbf{X}^\star$ from the (possibly noisy) observation of $|\mathbf{\Phi} \mathbf{X}^\star|$, in which $\mathbf{\Phi}$ is a matrix of size $m \times n$. We consider a \emph{high-dimensional} setting where $n,m \to \infty$ with $m/n = \mathcal{O}(1)$, and a large class of (possibly correlated) random matrices $\mathbf{\Phi}$ and observation channels. Spectral methods are a powerful tool to obtain approximate observations of the signal $\mathbf{X}^\star$ which can be then used as initialization for a subsequent algorithm, at a low computational cost. In this paper, we extend and unify previous results and approaches on spectral methods for the phase retrieval problem. More precisely, we combine the linearization of message-passing algorithms and the analysis of the \emph{Bethe Hessian}, a classical tool of statistical physics. Using this toolbox, we show how to derive optimal spectral methods for arbitrary channel noise and right-unitarily invariant matrix $\mathbf{\Phi}$, in an automated manner (i.e. with no optimization over any hyperparameter or preprocessing function)., Comment: 14 pages + references and appendix. v2: Version updated to match the one accepted at MSML 2021. v3: Adding a reference to a previous work mentioning marginal stability and its connection to Bayes-optimality

We study the problem of learning an unknown function using random feature models. Our main contribution is an exact asymptotic analysis of such learning problems with Gaussian data. Under mild regularity conditions for the feature matrix, we provide an exact characterization of the asymptotic training and generalization errors, valid in both the under-parameterized and over-parameterized regimes. The analysis presented in this paper holds for general families of feature matrices, activation functions, and convex loss functions. Numerical results validate our theoretical predictions, showing that our asymptotic findings are in excellent agreement with the actual performance of the considered learning problem, even in moderate dimensions. Moreover, they reveal an important role played by the regularization, the loss function and the activation function in the mitigation of the "double descent phenomenon" in learning.

Estimating a binary vector from noisy linear measurements is a prototypical problem for MIMO systems. A popular algorithm, called the box-relaxation decoder, estimates the target signal by solving a least squares problem with convex constraints. This paper shows that the performance of the algorithm, measured by the number of incorrectly-decoded bits, has a limiting Poisson law. This occurs when the sampling ratio and noise variance, two key parameters of the problem, follow certain scalings as the system dimension grows. Moreover, at a well-defined threshold, the probability of perfect recovery is shown to undergo a phase transition that can be characterized by the Gumbel distribution. Numerical simulations corroborate these theoretical predictions, showing that they match the actual performance of the algorithm even in moderate system dimensions.

In sparse linear regression, the SLOPE estimator generalizes LASSO by penalizing different coordinates of the estimate according to their magnitudes. In this paper, we present a precise performance characterization of SLOPE in the asymptotic regime where the number of unknown parameters grows in proportion to the number of observations. Our asymptotic characterization enables us to derive the fundamental limits of SLOPE in both estimation and variable selection settings. We also provide a computational feasible way to optimally design the regularizing sequences such that the fundamental limits are reached. In both settings, we show that the optimal design problem can be formulated as certain infinite-dimensional convex optimization problems, which have efficient and accurate finite-dimensional approximations. Numerical simulations verify all our asymptotic predictions. They demonstrate the superiority of our optimal regularizing sequences over other designs used in the existing literature.

Gravitational lensing brightens the images of quasars by a significant amount and provides a powerful probe of the physical properties of quasar population and their host galaxies. At z=6.51, recently discovered quasar J0439 is the most distant gravitationally lensed quasar known and the first such object detected at the epoch of reionization. Shallow HST/ACS observations reveal a compact lensing system, with Einstein ring radius of 0.2" and a high magnification factor of 51 for the central point source. J0439 is the brightest quasar observed at both optical and far-infrared wavelengths, indicating strong black hole (BH) accretion activities accompanied by a powerful starburst with an intrinsic star formation rate SFR >1000 solar mass per year. A shallow, low resolution ALMA image has resolved the lensed quasar host into a 1"-long arc. We propose to carry out deep HST/WFC3 F125W and F160W imaging to study the quasar host galaxy. It will detect and spatially resolved the rest-frame UV radiation from the host galaxy, allowing measurements of its unobscured star formation rate and size. The combination of the new HST observations with approved deep ALMA observations will measure the host galaxy SED, and, further aided by the extra resolution gain from lensing magnification, probe star formation and dust obscuration in its nuclear region. With additional multi- wavelength observations from X-ray through radio, J0439 will provide a unique laboratory to study BH/galaxy co-evolution at cosmic dawn in unprecedented detail. In particular, future JWST observations would probe the ISM kinematics in this starburst/quasar system within the BH sphere of influence.

A high fraction of the most luminous high-redshift quasars was predicted to be lensed due to magnification bias. We report the discovery of a strongly lensed ultraluminous quasar at z = 6.51, the first such object detected at the epoch of reionization and the brightest quasar yet known at z > 5. High-resolution HST imaging reveals a multiple imaged system with a maximum image separation 0.2", best explained by a model of three quasar images lensed by a low luminosity galaxy at z~0.7, with a magnification factor of ~50. This discovery suggests that a significant population of strongly lensed, high redshift quasars could have been missed by previous surveys, as standard color selection fails when the quasar color is contaminated by the lensing galaxy flux.

This write-up document is a summary of International workshop on the project for the extended hadron experimental facility which was held from March 26 to 28, 2018 at KEK Tokai Campus. This document is a collection of Letter Of Intents (LOIs) related to the proposed beam lines in the extended hadron experimental facility at J-PARC.

Phylogenetically informed trait comparisons across entire communities show promise in advancing community ecology. We use this approach to better understand the composition of a community of winter annual plants with multiple decades of monitoring and detailed morphological, phenological and physiological measurements. Previous research on this system revealed a physiological trade-off among dominant species that accurately predicts population and community dynamics. Here we expanded our investigation to 51 species, representing 96% of individual plants recorded over 30 years, and analysed trait relationships in the context of species abundance and phylogenetic relationships. We found that the functional-trait trade-off scales to the entire community, albeit with diminished strength. It is strongest for dominant species and weakens as progressively rarer species are included. The trade-off has been consistently expressed over three decades of environmental change despite some turnover in the identity of dominant species.

Objectives: The aim of this review is to evaluate the ocular safety of orthokeratology (OrthoK) treatment of myopia correction and retardation. Data Sources: Clinical studies published in English and Chinese were identified from MEDLINE, EMBASE CNKI, CQVIP, and WANFANG DATA (all from 1980 to April 2015). The reference lists of the studies and the Science Citation Index were also searched. Selection Criteria: Relevant clinical studies including case series, case reports, patient/practitioner surveys, retrospective and prospective cohort studies, and clinical trials were all included in the review. The material of OrthoK lenses was limited to gas-permeable lens. Main Results: This review incorporated a total of 170 publications, including 58 English and 112 Chinese literature. The risk of microbial keratitis in overnight OrthoK was similar to that of other overnight modality. The most common complication was corneal staining. Other clinically insignificant side effects included epithelial iron deposit, prominent fribrillary lines, and transient changes of corneal biomechanical properties. There was no long-term effect of OrthoK on corneal endothelium. Conclusions: There is sufficient evidence to suggest that OrthoK is a safe option for myopia correction and retardation. Long-term success of OrthoK treatment requires a combination of proper lens fitting, rigorous compliance to lens care regimen, good adherence to routine follow-ups, and timely treatment of complications.

In this work, we have fabricated a novel difunctional magnetic fluorescent nanohybrid (DMFN) for the determination of cadmium ions (Cd2+) in water samples, where the &ldquo, off-on&rdquo, model and &ldquo, ion-imprinting&rdquo, technique were incorporated simultaneously. The DMFN were composed of CdTe/CdS core-shell quantum dots (QD) covalently linked onto the surface of polystyrene magnetic microspheres (PMM) and characterized using ultraviolet-visible spectroscopy (UV-Vis), fluorescence spectroscopy, and transmission electron microscopy (TEM). Based on the favorable magnetic and fluorescent properties of the DMFN, the chemical etching of ethylene diamine tetraacetic acid (EDTA) at the surface produced specific Cd2+ recognition sites and quenched the red fluorescence of outer CdTe/CdS QD. Under optimal determination conditions, such as EDTA concentration, pH, and interfering ions, the working curve of determining Cd2+ was obtained, the equation was obtained Y = 34,759X + 254,894 (R = 0.9863) with a line range 0.05&ndash, 8 &mu, M, and the detection limit was 0.01 &mu, M. Results showed that synthesized magnetic fluorescent microspheres had high sensitivity, selectivity, and reusability in detection. Moreover, they have significant potential value in fields such as biomedicine, analytical chemistry, ion detection, and fluorescence labeling.

The influence of acid strength was evaluated toward the selectivity to propene on conversion of 1-pentene. For the catalytic cracking of 1-pentene, the main reaction pathways and the molar ratio of propene to ethene (P/E ratio) were controlled by acid strength with the appropriate amount of total acid sites. The results showed that the P/E ratio increased with decreasing amounts of strong acid sites, since the activation energies of individual reaction pathways were influenced by acid strength to a different extent. Strong acid sites could promote pathway I′ (C52– → C22– + C32–) and pathway II′-1 (C62– → C22– + C42–), while weak acid sites preferred pathway II′ (2C52– → C10+ → C42– + C62–) and pathway II′-2 (C62– → 2C32–), since pathways II′ and II′-2 underwent some energetically favorable routes (tertiary–secondary, secondary–secondary) of carbenium ion intermediates. By manipulation of the acid strength distribution on ZSM-5, the P/E ratio and selectivity of propene could be significantly improved, sugg...

In our recent theoretical work, we have provided a complete 1-D theory for restoring signals with finite rate of innovations (FRI) from sparse Fourier measurements using low-rank Fourier interpolation. In this work, we are expanding this theory for restoring common sparse signals where the singularity of the FRI signals is at the same position, while the unknown weights are different for each channel. We consider two situations: one for the same Fourier sampling pattern, which corresponds to the multiple measurement vector problems (MMV), and the other for different sampling patterns. We will derive performance guarantees: one for the algebraic guraantee and the other for probabilistic guarantee from randomly sampled Fourier measurements. We will discuss the relationship between the newly derived bounds and those of multiple signal classification (MUSIC) algorithm and single channel low-rank Fourier interpolation.

We analyze the dynamics of an online algorithm for independent component analysis in the high-dimensional scaling limit. As the ambient dimension tends to infinity, and with proper time scaling, we show that the time-varying joint empirical measure of the target feature vector and the estimates provided by the algorithm will converge weakly to a deterministic measured-valued process that can be characterized as the unique solution of a nonlinear PDE. Numerical solutions of this PDE, which involves two spatial variables and one time variable, can be efficiently obtained. These solutions provide detailed information about the performance of the ICA algorithm, as many practical performance metrics are functionals of the joint empirical measures. Numerical simulations show that our asymptotic analysis is accurate even for moderate dimensions. In addition to providing a tool for understanding the performance of the algorithm, our PDE analysis also provides useful insight. In particular, in the high-dimensional limit, the original coupled dynamics associated with the algorithm will be asymptotically "decoupled", with each coordinate independently solving a 1-D effective minimization problem via stochastic gradient descent. Exploiting this insight to design new algorithms for achieving optimal trade-offs between computational and statistical efficiency may prove an interesting line of future research., Comment: 10 pages, 3 figures, 31st Conference on Neural Information Processing Systems (NIPS 2017)

We study a spectral initialization method that serves a key role in recent work on estimating signals in nonconvex settings. Previous analysis of this method focuses on the phase retrieval problem and provides only performance bounds. In this paper, we consider arbitrary generalized linear sensing models and present a precise asymptotic characterization of the performance of the method in the high-dimensional limit. Our analysis also reveals a phase transition phenomenon that depends on the ratio between the number of samples and the signal dimension. When the ratio is below a minimum threshold, the estimates given by the spectral method are no better than random guesses drawn from a uniform distribution on the hypersphere, thus carrying no information; above a maximum threshold, the estimates become increasingly aligned with the target signal. The computational complexity of the method, as measured by the spectral gap, is also markedly different in the two phases. Worked examples and numerical results are provided to illustrate and verify the analytical predictions. In particular, simulations show that our asymptotic formulas provide accurate predictions for the actual performance of the spectral method even at moderate signal dimensions.

We present a framework for analyzing the exact dynamics of a class of online learning algorithms in the high-dimensional scaling limit. Our results are applied to two concrete examples: online regularized linear regression and principal component analysis. As the ambient dimension tends to infinity, and with proper time scaling, we show that the time-varying joint empirical measures of the target feature vector and its estimates provided by the algorithms will converge weakly to a deterministic measured-valued process that can be characterized as the unique solution of a nonlinear PDE. Numerical solutions of this PDE can be efficiently obtained. These solutions lead to precise predictions of the performance of the algorithms, as many practical performance metrics are linear functionals of the joint empirical measures. In addition to characterizing the dynamic performance of online learning algorithms, our asymptotic analysis also provides useful insights. In particular, in the high-dimensional limit, and due to exchangeability, the original coupled dynamics associated with the algorithms will be asymptotically "decoupled", with each coordinate independently solving a 1-D effective minimization problem via stochastic gradient descent. Exploiting this insight for nonconvex optimization problems may prove an interesting line of future research.

We consider a recently proposed convex formulation, known as the PhaseMax method, for solving the phase retrieval problem. Using the replica method from statistical mechanics, we analyze the performance of PhaseMax in the high-dimensional limit. Our analysis predicts the \emph{exact} asymptotic performance of PhaseMax. In particular, we show that a sharp phase transition phenomenon takes place, with a simple analytical formula characterizing the phase transition boundary. This result shows that the oversampling ratio required by existing performance bounds in the literature can be significantly reduced. Numerical results confirm the validity of our replica analysis, showing that the theoretical predictions are in excellent agreement with the actual performance of the algorithm, even for moderate signal dimensions.

Multiple scale-dependent ecological processes infl uence species distributions. Uncovering these drivers of dynamic range boundaries can provide fundamental ecological insights and vital knowledge for species management. We develop a transferable methodology that uses widely available data and tools to determine critical scales in range expansion and to infer dominating scale-dependent forces that infl uence spread. We divide a focal geographic region into diff erent sized square cells, representing diff erent spatial scales. We then used herbarium records to determine the species ’ occupancy of cells at each spatial scale. We calculated the growth in cell occupancy across scales to infer the scale dependent expansion rate. Th is is the fi rst time such a ‘ box-counting ’ method is used to study range expansion. We coupled this multi-scale analysis with species distribution models to determine the range and spatial scales where suitable climate allows the species to spread, and where other factors may be infl uencing the expansion. We demonstrate our methodology by assessing the spread of invasive Sahara mustard in North America. We detect critical scales where its spread is limited (100 – 500 km) or unconstrained (5 – 50 km) by climatic variables. Using climate-based models to assess the similarity of climate envelopes in its native and invaded range, we fi nd that the climate in the invaded range generally predicts the native distribution, suggesting that either there has been little local adaptation to climate occurring since introduction or the biological interaction experienced in the invaded range has not driven the species to occupy climatic conditions much diff erent from its native range. Our novel method can be broadly utilized in other studies to generate critical insights into the scale dependency of diff erent ecological drivers that infl uence the spread and distribution limits, as well as to help parameterizing predictions of future spread, and thus inform management decisions.

Nanoparticle research and development have brought significant breakthroughs in many areas of basic and applied sciences. However, efficiently collecting nanoparticles in large quantities in pure and natural systems is a major challenge in nanoscience. This review article has focused on experimental investigation and implications of nanoparticles in soil, clay, geological and environmental sciences. An automated ultrafiltration device (AUD) apparatus was used to demonstrate efficient collection and separation of nanoparticles in highly weathering red soils, black soils, and gouge of earthquake fault, as well as zeolite. The kaolinite, illite, goethite, and hematite were identified in highly weathering red soils. Transmission electron microscopic (TEM) images showed the presence of hematite nanoparticles on the surface coating of kaolinite nanoparticles and aggregated hematite nanoparticles overlapping the edge of a kaolinite flake in a size range from 4 to 7 nm. The maximum crystal violet (CV) and methylene blue (MB) adsorption amount of smectite nanoparticles (

Summary 1. The relationship between environmental productivity and the number of species [species richness–productivity relationship (SRPR)] has been thoroughly studied, but the mechanisms responsible for its form are still largely unknown, possibly because the majority of studies have focused on evaluating the sole effect of a single hypothesis. 2. We tested whether variation in species richness along a productivity gradient is due to variation in (i) the number of individuals, (ii) the number of species in the species pool or (iii) habitat heterogeneity. 3. We measured species richness (S), individual abundance (N) and productivity (P) estimated as standing biomass in different herbaceous communities in the Czech Republic at two spatial scales. Species pool (Spool) was obtained from a database concerning individual habitats, and habitat heterogeneity (H) was measured using the community dissimilarity index. 4. The SRPR was scale-dependent: at the smaller spatial scale of individual plots, there was a significant curvilinearly negative relationship between S and P, whereas at the larger site scale it turned into a non-significant relationship. 5. Species richness was significantly affected by a combined effect of N and Spool at the plot scale and by a combined effect of Spool and H at the site scale. None of these variables was sufficient to explain the SRPR by itself. 6. Synthesis. Our findings indicate that there is no universal form of the species–productivity relationship, and the SRPR is driven by multiple scale-dependent mechanisms. It is important to consider the joint effect of different factors in explaining species richness patterns rather than to focus on the sole effect of productivity.

We study the dynamics of an online algorithm for learning a sparse leading eigenvector from samples generated from a spiked covariance model. This algorithm combines the classical Oja's method for online PCA with an element-wise nonlinearity at each iteration to promote sparsity. In the high-dimensional limit, the joint empirical measure of the underlying sparse eigenvector and its estimate provided by the algorithm is shown to converge weakly to a deterministic, measure-valued process. This scaling limit is characterized as the unique solution of a nonlinear PDE, and it provides exact information regarding the asymptotic performance of the algorithm. For example, performance metrics such as the cosine similarity and the misclassification rate in sparse support recovery can be obtained by examining the limiting dynamics. A steady-state analysis of the nonlinear PDE also reveals an interesting phase transition phenomenon. Although our analysis is asymptotic in nature, numerical simulations show that the theoretical predictions are accurate for moderate signal dimensions., Comment: 5 pages

APOBEC3B (A3B), a cytidine deaminase, is known to play important roles in creating de novo genomic C-to-T mutations in cancers and contribute to induction of genomic instability. In this study, we evaluated the roles of A3B in the progression and metastasis of human hepatocellular carcinoma (HCC). Using whole-transcriptome and whole-exome sequencing and quantitative real-time PCR, we found that A3B was overexpressed in human HCCs and A3B expression was significantly correlated with the proportion of genomic C-to-A and G-to-T mutations. Upon clinicopathological correlation, higher A3B expression was associated with more aggressive tumor behavior. Wild-type A3B (wt-A3B) overexpression in HCC cells promoted cell proliferation, and cell migratory and invasive abilities in vitro, and tumorigenicity and metastasis in vivo. On the other hand, knockdown of A3B suppressed cell proliferation, migratory and invasive abilities of HCC cells with high endogenous A3B level. However, to our surprise, overexpression of A3B deaminase-dead double mutant (E68A/E255Q) led to similar results as wt-A3B in HCC both in vitro and in vivo. Furthermore, overexpression of wt-A3B and mutant A3B both enhanced cell cycle progression in HCC cells. Altogether, our data demonstrated a novel deaminase-independent role of A3B in contributing to HCC tumorigenesis and metastasis. Citation Format: Irene O L Ng, Wei Ma, Daniel WH Ho, Karen MF Sze, Yue M. Tsui, Lo K. Chan, Joyce MF Lee. APOBEC3B promotes hepatocarcinogenesis and metastasis through novel deaminase-independent activity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 4014.

Initiation of breast-feeding within 1 h after birth has been associated with reduced neonatal mortality in a rural Ghanaian population. In South Asia, however, breast-feeding patterns and low birth weight rates differ and this relationship has not been quantified. Data were collected during a community-based randomized trial of the impact of topical chlorhexidine antisepsis interventions on neonatal mortality and morbidity in southern Nepal. In-home visits were conducted on d 1-4, 6, 8, 10, 12, 14, 21, and 28 to collect longitudinal information on timing of initiation and pattern of breast-feeding. Multivariable regression modeling was used to estimate the association between death and breast-feeding initiation time. Analysis was based on 22,838 breast-fed newborns surviving to 48 h. Within 1 h of birth, 3.4% of infants were breast-fed and 56.6% were breast-fed within 24 h of birth. Partially breast-fed infants (72.6%) were at higher mortality risk [relative risk (RR) = 1.77; 95% CI = 1.32-2.39] than those exclusively breast-fed. There was a trend (P = 0.03) toward higher mortality with increasing delay in breast-feeding initiation. Mortality was higher among late (> or = 24 h) compared with early (< 24 h) initiators (RR = 1.41; 95% CI = 1.08-1.86) after adjustment for low birth weight, preterm birth, and other covariates. Improvements in breast-feeding practices in this setting may reduce neonatal mortality substantially. Approximately 7.7 and 19.1% of all neonatal deaths may be avoided with universal initiation of breast-feeding within the first day or hour of life, respectively. Community-based breast-feeding promotion programs should remain a priority, with renewed emphasis on early initiation in addition to exclusiveness and duration of breast-feeding.

We develop a matched signal detection (MSD) theory for signals with an intrinsic structure described by a weighted graph. Hypothesis tests are formulated under different signal models. In the simplest scenario, we assume that the signal is deterministic with noise in a subspace spanned by a subset of eigenvectors of the graph Laplacian. The conventional matched subspace detection can be easily extended to this case. Furthermore, we study signals with certain level of smoothness. The test turns out to be a weighted energy detector, when the noise variance is negligible. More generally, we presume that the signal follows a prior distribution, which could be learnt from training data. The test statistic is then the difference of signal variations on associated graph structures, if an Ising model is adopted. Effectiveness of the MSD on graph is evaluated both by simulation and real data. We apply it to the network classification problem of Alzheimer's disease (AD) particularly. The preliminary results demonstrate that our approach is able to exploit the sub-manifold structure of the data, and therefore achieve a better performance than the traditional principle component analysis (PCA).

Removal of 2-chlorophenol from water using rice-straw derived ash (RSDA) was evaluated in this study to compare with commercial activated carbon. RSDA was obtained by burning rice-straw at 400 °C and 700 °C for 1 h. This ash can provide a better adsorbent for 2-chlorophenol. The adsorption capacities of RSDA at 400 °C and 700 °C are 37 and 52 mg g⁻¹ at pH 4, respectively, and decrease to 9.0 and 40 mg g⁻¹ at pH 10. Adsorption of either neutral or anionic 2-cholorphenol by the RSDA are shown as L-shaped nonlinear isotherms, suggesting surface adsorption rather than partitioning is occurring. At higher-burning temperatures, the surface area, porosity, point of zero charge and aromaticity of the resultant RSDA increase, but the oxygen content and surface acidity decrease. The combined effects result in a higher 2-chlorophenol adsorption of RSDA at 700 °C, which shows a slight pH effect on the adsorption of 2-chlorophenol, due to the lower content of oxygen-containing functional groups. Oxygen-containing functional groups contribute to surface acidity and H-bonding sites for adsorbed water, which compromises the interaction between 2-chlorophenol and the adsorbents. Thus, it suggests that rice-straw derived carbon (RSDC) can be used as an effective low-cost substitute material for activated carbon for removal of chlorophenols from wastewater.

Based on the characteristics at different scales between signal and measure noise, this paper studied wavelet denoise method on ECNT (Eddy Current Nondestructive Detection). After analyzing the two current denoise methods using wavelet transform, such as Mallat multi-scaling edge detecting or thresholds method, authors introduced a new method. The result shows that this method is better than two others.

By conducting [0, 1] treatment to time consuming of logistics system network key links, and regarding the time consumed by manufacture, inspection, storage, assembling, packing and market as a kind of existent extent of the joint and the time consumed by materials handling, transportation and logistics information as the connection strength between joints in a generalized multi-directional fuzzy map, a generalized multi-directional fuzzy map model of logistics system networks is built. The mutual flow among network joints and the special form of generalized fuzzy matrix is analyzed. Finally, an example of model building is given.© (1997) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.