Search

Swertia cincta , a plant of the genus Swertia in Gentianceae, has "heat-clearing" and detoxifying effects that normalize the gallbladder function in the treatment of jaundice. Although numerous studies on Swertia cincta have been performed, the absorption and pharmacokinetic behaviors remain unclear. In this study, the compounds of Swertia cincta in serum, bile, feces, and urine of rats were analyzed using a ultra-high-performance liquid chromatography-tandem mass spectrometry. A total of 9 prototype components and 48 metabolites were detected in biological samples. Furthermore, we determined the main components absorbed in the blood of Swertia cincta and established a method for simultaneously determining these components (sweroside, swertiamarin, and gentiopicroside) in positive ionization mode within 6 min. The quantitative method was successfully applied for the multiple-component pharmacokinetic study of Swertia cincta. [ABSTRACT FROM AUTHOR]

This paper develops a weak Galerkin (WG) finite element method of arbitrary order for the steady incompressible Magnetohydrodynamics equations. The WG scheme uses piecewise polynomials of degrees k (k ≥ 1) , k , k − 1 and k − 1 respectively for the approximations of the velocity, the magnetic field, the pressure, and the magnetic pseudo-pressure in the interior of elements, and uses piecewise polynomials of degree k for their numerical traces on the interfaces of elements. The method is shown to yield globally divergence-free approximations of the velocity and magnetic fields. We give existence and uniqueness results for the discrete scheme and derive optimal a priori error estimates. We also present a convergent linearized iterative algorithm. Numerical experiments are provided to verify the obtained theoretical results. • We develop a weak Galerkin method of arbitrary order for the steady incompressible MHD equations. • The method is shown to yield globally divergence-free approximations of the velocity and magnetic fields. • We give existence and uniqueness results for the discrete scheme and derive optimal a priori error estimates. • We also present a convergent linearized iterative algorithm. • Numerical experiments confirm the theoretical results. [ABSTRACT FROM AUTHOR]

Proliferation of Industrial Internet has dramatically changed the way we live and work. It brings convenience to our society and sometimes requires real-time processing of dramatic data at the same time. However, traditional paradigm of computing on the center cloud can't always meet such requirement, for the non-negligible time delay of data transmission and communication. Edge computing is a novel computing paradigm proposed to resolve such a problem. As a promising technology, it extends computing from cloud center to the edge of network. Edge computing has the advantage of low latency to achieve a shorter response time, as well as potential to address the concerns of energy consuming, bandwidth burden and security issue. In this paper, we give a survey about edge computing from the aspect of methodologies, application scenarios and its role in Industrial Internet. Some open issues of edge computing are also introduced in this paper. At the end of the manuscript, a discussion about future direction is proposed. The shallow network algorithms such as broad learning system (BLS), which have achieved great improvement in computing efficiency, show an optimistic outlook in this area. We propose our conceive about future applications when shallow network methods like BLS are applied in edge computing and hope the paper will inspire research in relative directions. [ABSTRACT FROM AUTHOR]

Purpose: The purpose of this paper is to propose three iterative finite element methods for equations of thermally coupled incompressible magneto-hydrodynamics (MHD) on 2D/3D bounded domain. The detailed theoretical analysis and some numerical results are presented. The main results show that the Stokes iterative method has the strictest restrictions on the physical parameters, and the Newton's iterative method has the higher accuracy and the Oseen iterative method is stable unconditionally. Design/methodology/approach: Three iterative finite element methods have been designed for the thermally coupled incompressible MHD flow on 2D/3D bounded domain. The Oseen iterative scheme includes solving a linearized steady MHD and Oseen equations; unconditional stability and optimal error estimates of numerical approximations at each iterative step are established under the uniqueness condition. Stability and convergence of numerical solutions in Newton and Stokes' iterative schemes are also analyzed under some strong uniqueness conditions. Findings: This work was supported by the NSF of China (No. 11971152). Originality/value: This paper presents the best choice for solving the steady thermally coupled MHD equations with different physical parameters. [ABSTRACT FROM AUTHOR]

Purpose: This paper aims to further develop stator permanent magnet (PM) type memory machines by providing generalized design guidelines for double-stator memory machines (DSMMs) with hybrid PMs. This paper discusses the design experience of DSMMs and presents a comparative study of radial magnetization (RM) and circumferential magnetization (CM) types. Design/methodology/approach: It begins with an introduction to RM and CM operating principles and magnetization mechanisms. Then, a comparative study is conducted for one of the RM-DSMM rotor pole pairs, inner and outer stator clamping angles and low coercive force PMs thickness. Finally, the two machines' finite element simulation performance is compared. The validity of the proposed machine structure is demonstrated. Findings: In this paper, the double-stator structure is extended to parallel hybrid PM memory machines, and two novel DSMMs with RM and CM configurations are proposed. Two types of DSMMs have PMs and magnetizing windings on the inner stator and armature windings on the outer stator. The main difference between the two is the arrangement of PMs on the inner stator. Originality/value: Conventional stator PM memory machines have geometrical space conflicts between the PM and armature windings. The proposed double-stator structure can alleviate these conflicts and increase the torque density accordingly. In addition, this paper contributes to comparing the arrangement of hybrid PMs for DSMMs. [ABSTRACT FROM AUTHOR]

A major progress in deep multilayer neural networks (DNNs) is the invention of various unsupervised pretraining methods to initialize network parameters which lead to good prediction accuracy. This paper presents the sparseness analysis on the hidden unit in the pretraining process. In particular, we use the L1 -norm to measure sparseness and provide some sufficient conditions for that pretraining leads to sparseness with respect to the popular pretraining models—such as denoising autoencoders (DAEs) and restricted Boltzmann machines (RBMs). Our experimental results demonstrate that when the sufficient conditions are satisfied, the pretraining models lead to sparseness. Our experiments also reveal that when using the sigmoid activation functions, pretraining plays an important sparseness role in DNNs with sigmoid (Dsigm), and when using the rectifier linear unit (ReLU) activation functions, pretraining becomes less effective for DNNs with ReLU (Drelu). Luckily, Drelu can reach a higher recognition accuracy than DNNs with pretraining (DAEs and RBMs), as it can capture the main benefit (such as sparseness-encouraging) of pretraining in Dsigm. However, ReLU is not adapted to the different firing rates in biological neurons, because the firing rate actually changes along with the varying membrane resistances. To address this problem, we further propose a family of rectifier piecewise linear units (RePLUs) to fit the different firing rates. The experimental results show that the performance of RePLU is better than ReLU, and is comparable with those with some pretraining techniques, such as RBMs and DAEs. [ABSTRACT FROM PUBLISHER]

Purpose: The purpose of this paper is to design a simple and accurate a-posteriori Lagrangian-based error estimator is developed for the class of backward differentiation formula (BDF) algorithms with variable time step size, and the adaptive time-stepping in BDF algorithms is demonstrated for efficient time-dependent simulations in fluid flow and heat transfer. Design/methodology/approach: The Lagrange interpolation polynomial is used to predict the time derivative, and then the accurate primary result is obtained by the Gauss integral, which is applied to evaluate the local error. Not only the generalized formula of the proposed error estimator is presented but also the specific expression for the widely applied BDF1/2/3 is illustrated. Two essential executable MATLAB functions to implement the proposed error estimator are appended for practical applications. Then, the adaptive time-stepping is demonstrated based on the newly proposed error estimator for BDF algorithms. Findings: The validation tests show that the newly proposed error estimator is accurate such that the effectivity index is always close to unity for both linear and nonlinear problems, and it avoids under/overestimation of the exact local error. The applications for fluid dynamics and coupled fluid flow and heat transfer problems depict the advantage of adaptive time-stepping based on the proposed error estimator for time-dependent simulations. Originality/value: In contrast to existing error estimators for BDF algorithms, the present work is more accurate for the local error estimation, and it can be readily extended to practical applications in engineering with a few changes to existing codes, contributing to efficient time-dependent simulations in fluid flow and heat transfer. [ABSTRACT FROM AUTHOR]

In deep learning, auxiliary modules for model training have become increasingly popular, such as Deep Mutual Learning (DML) and Multi-Scale Dense Convolutional Networks (MSDNet), which can maximize the performance of the model without increasing the amount of test computation. Nevertheless, current research does not fully exploit the knowledge between different auxiliary modules. This paper proposes a new model training technique–Siamese Labels Auxiliary (SiLA) Learning, in which the SiLA module is designed to concatenate the outputs of the modules to get auxiliary information. In addition, the knowledge learned from different modules can be fully exploited by auxiliary information. Our experiments show that SiLA Learning effectively improves the performance of standard models and achieves convincing experimental results on image classification tasks without increasing the amount of test computation. Moreover, the experiments show that SiLA Learning can be easily combined with methods such as DML and MSDNet to exploit the knowledge fully and perform best. [ABSTRACT FROM AUTHOR]

Ferroelectric materials often experience fracture and dielectric breakdown under large mechanical and/or electrical loads. These two failure behaviors can affect each other and may lead to new phenomena. This study predicts a new type of jumping dielectric breakdown during crack propagation in ferroelectric materials using a phase field model involving multiple order parameters. The driving forces for the evolution of polarization, fracture, and dielectric breakdown were analysed based on the concept of configurational forces. The jumping dielectric breakdown was attributed to the competition between crack propagation and breakdown initiation at the moving crack tip. The breakdown pattern is related to the polarization domain structure and is significantly influenced by the applied electric field and crack boundary conditions. Furthermore, the jumping breakdown associated with possible mechanical degradation of the material can change the driving and resistance forces for crack propagation. This study introduces a novel dielectric breakdown behavior during crack propagation and provides new insights into the coupling failure of ferroelectric materials. [ABSTRACT FROM AUTHOR]

Purpose: The purpose of this paper is to describe a novel universal error estimator and the adaptive time-stepping process in the generalized single-step single-solve (GS4-1) computational framework, applied for the fluid dynamics with illustrations to incompressible Navier–Stokes equations. Design/methodology/approach: The proposed error estimator is universal and versatile that it works for the entire subsets of the GS4-1 framework, encompassing the nondissipative Crank–Nicolson method, the most dissipative backward differential formula and anything in between. It is new and novel that the cumbersome design work of error estimation for specific time integration algorithms can be avoided. Regarding the numerical implementation, the local error estimation has a compact representation that it is determined by the time derivative variables at four successive time levels and only involves vector operations, which is simple for numerical implementation. Additionally, the adaptive time-stepping is further illustrated by the proposed error estimator and is used to solve the benchmark problems of lid-driven cavity and flow past a cylinder. Findings: The proposed computational procedure is capable of eliminating the nonphysical oscillations in GS4-1(1,1)/Crank–Nicolson method; being CPU-efficient in both dissipative and nondissipative schemes with better solution accuracy; and detecting the complex physics and hence selecting a suitable time step according to the user-defined error threshold. Originality/value: To the best of the authors' knowledge, for the first time, this study applies the general purpose GS4-1 family of time integration algorithms for transient simulations of incompressible Navier–Stokes equations in fluid dynamics with constant and adaptive time steps via a novel and universal error estimator. The proposed computational framework is simple for numerical implementation and the time step selection based on the proposed error estimation is efficient, benefiting to the computational expense for transient simulations. [ABSTRACT FROM AUTHOR]

• The proposed multiscale framework effectively connects quantum, nano-, micro- and macroscale phenomena. • Surface eigenstress model combined with symbolic regression to obtain constitutive equations for finite element calculation. • Generative design method for mapping electrochemical actuator structures using 0 s and 1 s number sequences. • Surrogate model improves optimization efficiency by at least 5 orders of magnitude over using finite element calculations. • Pareto optimization theory for obtaining structures with both large actuation strain and high effective Young's modulus. Nanoporous materials (NMs) immersed in electrolytes can achieve approximately 1 % deformation at a low operating voltage of about 1 V. The actuation renders them promising artificial muscles. The actuation performance significantly hinges on the structure and size of nanopores and ligaments in NMs. Consequently, designing an optimal configuration is imperative for excellent performance. The actuation mechanism of NMs involves the coupling of multiple fields at various length scales, posing a formidable challenge to conventional simulation and design approaches. To surmount this challenge, we have developed a computational framework capable of conducting concurrent and sequential multiscale calculations. By utilizing artificial neural network (ANN) surrogate models trained on data obtained through the finite element method (FEM), the framework achieves optimized values for both actuation strain and effective Young's modulus within a designated design space. The constitutive model, which establishes the relationship between surface stress and charges in FEM, is derived from the surface eigenstress model and symbolic regression. This involves utilizing data calculated through joint density functional theory. This framework not only ensures the desired properties but also demonstrates its potential for effectively addressing other multiscale optimization problems. [Display omitted] [ABSTRACT FROM AUTHOR]

Video object segmentation (VOS) is one of the most fundamental tasks for numerous sequent video applications. The crucial issue of online VOS is the drifting of segmenter when incrementally updated on continuous video frames under unconfident supervision constraints. In this work, we propose a self-teaching VOS (ST-VOS) method to make segmenter to learn online adaptation confidently as much as possible. In the segmenter learning at each time slice, the segment hypothesis and segmenter update are enclosed into a self-looping optimization circle such that they can be mutually improved for each other. To reduce error accumulation of the self-looping process, we specifically introduce a metalearning strategy to learn how to do this optimization within only a few iteration steps. To this end, the learning rates of segmenter are adaptively derived through metaoptimization in the channel space of convolutional kernels. Furthermore, to better launch the self-looping process, we calculate an initial mask map through part detectors and motion flow to well-establish a foundation for subsequent refinement, which could result in the robustness of the segmenter update. Extensive experiments demonstrate that this ST idea can boost the performance of baselines, and in the meantime, our ST-VOS achieves encouraging performance on the DAVIS16, Youtube-objects, DAVIS17, and SegTrackV2 data sets, where, in particular, the accuracy of 75.7% in J-mean metric is obtained on the multi-instance DAVIS17 data set. [ABSTRACT FROM AUTHOR]

Despite the increasing interest in multiagent reinforcement learning (MARL) in multiple communities, understanding its theoretical foundation has long been recognized as a challenging problem. In this article, we address this problem by providing a finite-sample analysis for decentralized batch MARL. Specifically, we consider a type of mixed MARL setting with both cooperative and competitive agents, where two teams of agents compete in a zero-sum game setting, while the agents within each team collaborate by communicating over a time-varying network. This setting covers many conventional MARL settings in the literature. We then develop batch MARL algorithms that can be implemented in a decentralized fashion, and quantify the finite-sample errors of the estimated action-value functions. Our error analysis captures how the function class, the number of samples within each iteration, and the number of iterations determine the statistical accuracy of the proposed algorithms. Our results, compared to the finite-sample bounds for single-agent reinforcement learning, involve additional error terms caused by decentralized computation, which is inherent in our decentralized MARL setting. This article provides the first finite-sample analysis for batch MARL, a step toward rigorous theoretical understanding of general MARL algorithms in the finite-sample regime. [ABSTRACT FROM AUTHOR]

Electronic medical records (EMRs) play an important role in medical data mining and sequential data learning. In this article, we propose to use a sequential neural network with dynamic content-based memories to predict future medications, given EMRs. The local–global memory neural network contains two layers of memories: the local memory and the global memory. Particularly, our method learns the hidden knowledge within EMRs by locally remembering individual patterns of a patient (via local memory) and globally remembering group evidence of disease (via global memory). In addition, we show how our model can be modified to classify the hidden states of EMRs from different patients at each time step into different phases that indicate the progressions of medications in terms of a specific disease, in an unsupervised manner. Experimental results on real EMRs data sets show that, by learning EMRs with external local and global memories, with regard to a given disease, our model improves the prediction performance compared with several alternative methods. [ABSTRACT FROM AUTHOR]

Graph classification is a fundamental but challenging issue for numerous real-world applications. Despite recent great progress in image/video classification, convolutional neural networks (CNNs) cannot yet cater to graphs well because of graphical non-Euclidean topology. In this article, we propose a walk-steered convolutional (WSC) network to assemble the essential success of standard CNNs, as well as the powerful representation ability of random walk. Instead of deterministic neighbor searching used in previous graphical CNNs, we construct multiscale walk fields (a.k.a. local receptive fields) with random walk paths to depict subgraph structures and advocate graph scalability. To express the internal variations of a walk field, Gaussian mixture models are introduced to encode the principal components of walk paths therein. As an analogy to a standard convolution kernel on image, Gaussian models implicitly coordinate those unordered vertices/nodes and edges in a local receptive field after projecting to the gradient space of Gaussian parameters. We further stack graph coarsening upon Gaussian encoding by using dynamic clustering, such that high-level semantics of graph can be well learned like the conventional pooling on image. The experimental results on several public data sets demonstrate the superiority of our proposed WSC method over many state of the arts for graph classification. [ABSTRACT FROM AUTHOR]

Decision-theoretic rough sets (DTRSs), which can be considered as generalized rough set models produced by Bayesian risk minimum and three-way decisions (3WD) theories, have achieved fruitful results in risk decision-making problems. Nevertheless, the parameter determination of decision-theoretic rough sets is a challenging problem in practical applications, which narrows the generalization and development of these models. In this paper, a methodology to determine the parameters for DTRS and 3WD is proposed to improve their practicability. First, a data-driven loss function matrix is introduced based on the significance and the probability of the sample. Subsequently, a generalized rough set model named single-parameter decision-theoretic rough set (SPDTRS) is put forward based on the proposed loss function matrix. The main feature of the proposed model is that there is only one parameter that should be preset rather than the two or six parameters in the traditional DTRS models. Finally, some experiments on the University of California Irvine (UCI) and Knowledge Extraction based on Evolutionary Learning (KEEL) data sets are conducted to illustrate the effectiveness of the proposed methodology. [ABSTRACT FROM AUTHOR]

Fisher’s criterion is one of the most popular discriminant criteria for feature extraction. It is defined as the generalized Rayleigh quotient of the between-class scatter distance to the within-class scatter distance. Consequently, Fisher’s criterion does not take advantage of the discriminant information in the class covariance differences, and hence, its discriminant ability largely depends on the class mean differences. If the class mean distances are relatively large compared with the within-class scatter distance, Fisher’s criterion-based discriminant analysis methods may achieve a good discriminant performance. Otherwise, it may not deliver good results. Moreover, we observe that the between-class distance of Fisher’s criterion is based on the $\ell _{2}$ -norm, which would be disadvantageous to separate the classes with smaller class mean distances. To overcome the drawback of Fisher’s criterion, in this paper, we first derive a new discriminant criterion, expressed as a mixture of absolute generalized Rayleigh quotients, based on a Bayes error upper bound estimation, where mixture of Gaussians is adopted to approximate the real distribution of data samples. Then, the criterion is further modified by replacing $\ell _{2}$ -norm with $\ell _{1}$ one to better describe the between-class scatter distance, such that it would be more effective to separate the different classes. Moreover, we propose a novel $\ell _{1}$ -norm heteroscedastic discriminant analysis method based on the new discriminant analysis (L1-HDA/GM) for heteroscedastic feature extraction, in which the optimization problem of L1-HDA/GM can be efficiently solved by using the eigenvalue decomposition approach. Finally, we conduct extensive experiments on four real data sets and demonstrate that the proposed method achieves much competitive results compared with the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

In the present work, we performed experiments, atomistic simulations, and high-resolution electron microscopy (HREM) to study the creep behaviors of the nanotwinned (nt) and nanograined (ng) copper at temperatures of 22 °C (RT), 40 °C, 50 °C, 60 °C, and 70 °C. The experimental data at various temperatures and different sustained stress levels provide sufficient information, which allows one to extract the deformation parameters reliably. The determined activation parameters and microscopic observations indicate transition of creep mechanisms with variation in stress level in the nt-Cu, i.e., from the Coble creep to the twin boundary (TB) migration and eventually to the perfect dislocation nucleation and activities. The experimental and simulation results imply that nanotwinning could be an effective approach to enhance the creep resistance of twin-free ng-Cu. The experimental creep results further verify the newly developed formula (Yang et al., 2016) that describes the time-, stress-, and temperature-dependent plastic deformation in polycrystalline copper. [ABSTRACT FROM AUTHOR]