基于WEB方式的电子政务初探

本提出了基于WEB方式的电子政务设计、开发、运营方法和技术平台的总体框架。描述了电子政务数据开发模型的雄度，电子政务系统工程的实施原则、实施步骤及应用系统集成的一整套的实践模型。     

Nonlinear Time Series Analysis Since 1990:Some Personal Reflections

Abstract I reflect upon the development of nonlinear time series analysis since 1990 by focusing on five majorareas of development. These areas include the interface between nonlinear time series analysis and chaos,thenonparametric/semiparametric approach,nonlinear state space modelling,financial time series and nonlinearmodelling of panels of time series.     

Protective Properties of Zinc-Rich Coatings: An Impedance Method Estimate

The impedance of zinc-rich polymer coatings on steel in a 3-% NaCl solution is studied. The electrochemical behavior of the electrode is satisfactorily described by an equivalent circuit that contains a constant-phase element, which reflects fractal properties of the zinc surface in the binder. From the results of calculations of the circuit elements for metal-rich electrodes with different zinc contents computed are fractions of the active area of the surface of the coating and its fractal dimensionality.     

Intermolecular hydrogen bonds: Comparison of associates in crystals and liquids

The density and sound velocity were measured within wide ranges of temperature for a number of liquids (water, formamide, diols, aliphatic alcohols, cellosolves, ketones), in which various types of H-associate can exist. The temperature dependences of the adiabatic and molar adiabatic compressibility, density, and molar volume are analyzed. The values of the molar adiabatic compressibility permit one to evaluate the dimensionality of H-associates existing in liquids; the values of adiabatic compressibility do not offer this possibility. The terms responsible for the similarity and difference between H-associates in crystals and liquids are discussed.     

Monte—Carlo model of the diffusion-limited reaction kinetics in microheterogeneous media. Comparison with the experimental plots of the forward and backward electron phototransfer kinetics on the pore diameter in silica gel

The kinetics of the diffusion-limited decay reaction A + B B was simulated by the Monte—Carlo method on a two-dimensional square lattice with defects presented by randomly distributed sites. The cases were considered where [B] [A] at the random initial distribution (quenching reaction) and [B] = [A] with the initial distribution of the A and B particles on neighboring sites (geminate recombination). The kinetic curves were approximated by the simplest analytical equation [A]/[A]₀ = (1 – )exp[–(kt)^1–h ] + (where k and are constants). The plots of the heterogeneity parameter (h) and time-averaged first-order rate constant vs. concentration of defects (p) or B particles (in the case of quenching) were obtained and compared with similar correlations obtained earlier by the experimental study of the kinetics of forward (quenching reaction) and backward (geminate recombination) electron phototransfer on the surface of different porous silica gels. The experimental plots of h vs. silica gel porosity are in satisfactory agreement with the plots of h vs. p in the model space, if the fraction of volume inaccessible for reactants, calculated from the free silica gel volume, is chosen as the p parameter for silica gel.Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 8, pp. 1536–1541, August, 2004.     

一种新的彩色图像降维方法

基于内容的图像检索(CBIR)是图像检索的重要分支，而基于颜色的特征提取是CBIR的常用方法之一．如果对图像颜色的特征数提取过多、维数过大，则不利于对图像的快速匹配．本文将图像的色彩直方图作为输入向量，然后采用局部线性映射(LLE)算法对原始数据进行降维，并分别在4种色彩空间下对降维后的彩色图像进行分类．实验证明，在处理非线性数据降维时，LLE较主成分分析(PCA)具有明显的优势．     

基于属性降维的钓鱼网站检测方法

随着对钓鱼网站分析刻画的不断完善,使得钓鱼网站检测特征呈现高维化特点。属性维度的增加以及数据量的增长,会造成检测计算复杂度呈几何倍数扩大,导致检测时间复杂度高、占用资源大和检测效率低。针对多属性的钓鱼网站检测,设计了一种基于属性降维的钓鱼网站检测方法。该方法使用信息增益方法对原始数据进行特征选择,筛除可能存在的冗余和噪声数据信息;根据不同属性间的互信息计算属性相关性矩阵,利用属性相关性矩阵作为权值参与加权主成分分析;根据得到的降维后新特征数据通过监督学习算法构建钓鱼网站检测模型。实验表明,该方法可以有效降低原始数据中冗余和噪声属性的干扰,能够有效检测出复杂网络环境中的钓鱼网站,同时具有较高的稳定性。     

Local Intrinsic Dimensionality,Entropy and Statistical Divergences

Properties of data distributions can be assessed at both global and local scales. At a highly localized scale, a fundamental measure is the local intrinsic dimensionality (LID), which assesses growth rates of the cumulative distribution function within a restricted neighborhood and characterizes properties of the geometry of a local neighborhood. In this paper, we explore the connection of LID to other well known measures for complexity assessment and comparison, namely, entropy and statistical distances or divergences. In an asymptotic context, we develop analytical new expressions for these quantities in terms of LID. This reveals the fundamental nature of LID as a building block for characterizing and comparing data distributions, opening the door to new methods for distributional analysis at a local scale.     

A Preprocessing Perspective for Quantum Machine Learning Classification Advantage in Finance Using NISQ Algorithms

Quantum Machine Learning (QML) has not yet demonstrated extensively and clearly its advantages compared to the classical machine learning approach. So far, there are only specific cases where some quantum-inspired techniques have achieved small incremental advantages, and a few experimental cases in hybrid quantum computing are promising, considering a mid-term future (not taking into account the achievements purely associated with optimization using quantum-classical algorithms). The current quantum computers are noisy and have few qubits to test, making it difficult to demonstrate the current and potential quantum advantage of QML methods. This study shows that we can achieve better classical encoding and performance of quantum classifiers by using Linear Discriminant Analysis (LDA) during the data preprocessing step. As a result, the Variational Quantum Algorithm (VQA) shows a gain of performance in balanced accuracy with the LDA technique and outperforms baseline classical classifiers.     

基于DDR的字典学习在光照变化人脸识别中的应用

针对光照变化人脸识别中大多数现有的人脸识别算法只能单独实施降维或者字典学习而不能完全利用训练样本判别信息的问题,提出了基于判别性降维的字典学习算法。首先,利用经典的特征提取算法PCA初始化降维投影矩阵;然后,计算字典和系数,通过联合降维与字典学习使得投影矩阵和字典更好地相互拟合;最后,利用迭代算法输出字典和投影矩阵,并利用经l2-范数正则化的分类器完成人脸的识别。在PIE及扩展的YaleB两大人脸数据库上得到了验证了所提算法的有效性及可靠性,实验结果表明,相比几种较为先进的线性表示算法,所提算法在处理光照变化人脸识别时取得了更高的识别率。