A tutorial on support vector machine-based methods for classification problems in chemometrics

This tutorial provides a concise overview of support vector machines and different closely related techniques for pattern classification. The tutorial starts with the formulation of support vector machines for classification. The method of least squares support vector machines is explained. Approaches to retrieve a probabilistic interpretation are covered and it is explained how the binary classification techniques can be extended to multi-class methods. Kernel logistic regression, which is closely related to iteratively weighted least squares support vector machines, is discussed. Different practical aspects of these methods are addressed: the issue of feature selection, parameter tuning, unbalanced data sets, model evaluation and statistical comparison. The different concepts are illustrated on three real-life applications in the field of metabolomics, genetics and proteomics.     

带Brown运动的随机奇异积分的存在性

在轨迹二阶导数具有Hlder连续的条件下,利用高阶奇异积分思想和概率极限的理论,研究了在Brown运动下的随机奇异积分.得到了以Brown运动为积分元的随机奇异积分是存在性定理.     

Operational and design considerations for broad area graded barrier quantum well heterostructure lasers grown by metalorganic chemical vapor deposition for high power applications

Graded barrier quantum well heterostructure (GBQWH) broad area lasers have been shown to be capable of high power pulsed and cw operation. In this article, we consider several operational characteristics and design issues associated with broad area graded barrier quantum well heterostructure lasers grown by metalorganic chemical vapor deposition. In particular, the effect of junction heating on emission wavelength for cw device operation and the effects of various buffer layer structures on the material properties and device characteristics of GBQWH structures are addressed. Typical results for high power operation of uncoated broad area laser diodes are also outlined.     

大型稀疏线性互补问题的行作用法

且引言考虑线性互补问题＊＊P（q，M）：求X二（X；，x。，…，x。厂E”使得x＞O，训x）E＊x＋g＞o，／U（X）一O（1）其中M一（m；。）为nXn矩阵（不必对称），q一切，q。，…，q。）rER“为给定常向量．通常情况下已有求解LCP（q，M）的若干著名算法［‘－’j．本文提出求解LCP（q，M）的一种新算法一行作用法，方法具有如下特点：（i）每次迭代只需n个简单的投影运算，每次投影只涉及矩阵M的一行；（n）生成新的迭代点x‘“‘时只利用前次迭代点／；（iii）对矩阵M不实施任何整体运算．因而适合于求解大型（巨型）稀疏问题，且…     

结构优化的RBF神经网络学习算法

文章提出了一种自动“删减”隐层神经元的ＲＢＦ神经网络学习算法。模拟结果表明，该算法训练的ＲＢＦ网络不仅结构得以优化，同时性能良好，可能成功地应用于模式分类和时间序列预测问题中。     

The use of dynamic basis functions in proper orthogonal decomposition

Currently, the usefulness of proper orthogonal decomposition (POD) is limited to computational domains with fixed meshes and fixed boundaries. This paper presents a new POD method that enables the modeling of flow through computational domains with deforming meshes and/or moving boundaries. To achieve this goal, the solution is approximated using basis functions which, although not explicitly functions of time, depend on parameters associated with flow unsteadiness. Results are shown for transonic flow through the Tenth Standard Configuration. Comparisons are made between this method and the standard approach for on- and off-reference flow conditions. This method properly captured flow nonlinearities and shock motion for cases in which the classical POD method failed.     

Transport property characterization of sparse CNT networks via AFM images

A simulation method for correlating the resistivity and resistance of sparse carbon nanotube (CNT) networks via atomic force microscopy images was proposed. For the demonstration, resistance values simulated by this method were compared with values obtained by directory measuring the resistance of sparse CNT networks. Results were also compared with those obtained by a thin-film approximation in which CNT networks are approximated as thin rectangles. Simulated resistance values were closer to the experimental values of the same samples than those estimated on the basis of the thin-film approximation. The use of atomic force microscopy (AFM) images enabled the implementation of inhomogeneity to numerical models, as well as one-to-one comparison between real samples and numerical models.     

Chemical profiling of Qingfei Paidu Decoction by triplex off-line two-dimensional liquid chromatography coupled with quadrupole time-of-flight mass spectrometry

Qingfei Paidu Decoction is a Chinese medicine formula that has been proved effective in the treatment of coronavirus disease 2019. However, the comprehensive separation and characterization of Qingfei Paidu Decoction are of a great challenge due to the diversity of chemical components in a wide range of polarity. In this study, a triplex off-line two-dimensional liquid chromatography coupled with quadrupole time-of-flight mass spectrometry is developed for the analysis of Qingfei Paidu Decoction. One reversed-phase liquid chromatography×hydrophilic interaction liquid chromatography system and two reversed-phase liquid chromatography×reversed phase liquid chromatography systems were constructed to separate polar components and weak-polar components in Qingfei Paidu Decoction, respectively. Benefiting from the good orthogonality of two-dimensional liquid chromatography and high sensitivity of quadrupole time-of-flight MS, chemical components with different polarities and content were discovered. A total of 749 peaks were detected in positive and negative ionization mode and presented as a four-dimensional data plot. Meanwhile, 498 compounds belonging to 14 categories were tentatively identified. These results provide good supplementary to elucidate the material basis of Qingfei Paidu Decoction. The triplex off-line two-dimensional liquid chromatography-quadrupole time-of-flight mass spectrometry strategy can be a powerful and efficient tool for the separation and characterization of chemical substances in traditional Chinese medicine formulas.     

Identification of Pipeline Inner Wall Geometry Based on the POD⁃RBF Method北大核心CSCD

Based on the proper orthogonal decomposition⁃radial basis function (POD⁃RBF), a geometric identification method for pipeline inner wall was proposed to solve the internal corrosion detection problem of natural gas and oil pipelines. In view of the static magnetic field, the simplified finite element model for the pipelines was established, and the variable⁃geometry sample library was constructed, to realize the response prediction of arbitrary geometry by the POD⁃RBF. The proposed method achieves reduced⁃order analysis and avoids repeated solution of the stiffness matrix due to the geometrical change during the identification process. Hence, it can significantly improve the computation efficiency. Finally, the grey wolf optimization (GWO) algorithm was used to optimize the objective function and avoid the calculation of the sensitivity in the process of geometry change. The numerical examples show that, the proposed method has high efficiency and accuracy in the geometric identification of the pipeline inner wall, with good stability even under introduced noises. © 2023 Editorial Office of Applied Mathematics and Mechanics. All rights reserved.     

基于压缩感知的实时目标追踪算法

提出了一种高效的基于压缩感知的实时目标追踪算法,该算法将空域数据转换到小波域,然后利用变密度采样矩阵对小波域数据进行压缩,从而极大地降低了数据量。在稀疏重建上,将St-OMP算法代替OMP算法以提高稀疏重建的速度。在多种具有挑战性的视频序列上进行实验,结果表明该算法提高了追踪准确度和速度。