On compressibility of osmium metal

On the basis of the high-pressure diamond anvil cell experiments on Os metal, Cynn et al. [Phys. Rev. Lett. 88, 135701-1 (2002)] have reported that this metal has lower compressibility than diamond. In the present work we have reanalysed the experimental data of Cynn et al. We find that the bulk moduli of Os and diamond are close to each other, implying that Os metal is as incompressible as diamond, but not more so. Our first principles total energy calculations using the full potential linearised augmented plane wave method on Os and diamond also suggest the same results.     

用多元分析方法研究乳腺癌血清的表面增强拉曼光谱

借用两种化学计量学分析方法——主成分分析方法(PCA)和判别分析方法(LDA),进行联合运用,分析了乳腺癌血清的表面增强拉曼光谱(SERS)。首先从分子角度上讨论了蛋白质主侧链在乳腺癌血清中变化,结果发现:主链空间结构遭到破坏。侧链的有序性下降、糖质环境也发生了改变;其次采用上述两种计量学的多元分析方法对18例临床确诊为乳腺癌患者和20例健康人的血清样本进行了表面增强拉曼光谱分析。发现主成分分析方法对于样本区分未能达到80%,进而再用判别分析法,准确率达到100%,所得结果令人满意。     

Multiclass Probability Estimation With Support Vector Machines

Multiclass classification and probability estimation have important applications in data analytics. Support vector machines (SVMs) have shown great success in various real-world problems due to their high classification accuracy. However, one main limitation of standard SVMs is that they do not provide class probability estimates, and thus fail to offer uncertainty measure about class prediction. In this article, we propose a simple yet effective framework to endow kernel SVMs with the feature of multiclass probability estimation. The new probability estimator does not rely on any parametric assumption on the data distribution, therefore, it is flexible and robust. Theoretically, we show that the proposed estimator is asymptotically consistent. Computationally, the new procedure can be conveniently implemented using standard SVM softwares. Our extensive numerical studies demonstrate competitive performance of the new estimator when compared with existing methods such as multiple logistic regression, linear discrimination analysis, tree-based methods, and random forest, under various classification settings. Supplementary materials for this article are available online.     

基于概率主题模型的文档聚类

 为了实现普通文本语料库和数字图书语料库的有效聚类,分别提出基于传统LDA(Latent Dirichlet Allocation)模型和TC_LDA模型的聚类算法.TC_LDA模型在LDA模型基础上进行扩展,通过对图书文档的目录和正文信息联合进行主题建模.和传统方法不同,基于主题模型的聚类算法能将具备同一主题的文档聚为一类.实验结果表明从主题分析角度出发实现的聚类算法优于传统的聚类算法.     

First principles calculations of Cd and Zn chalcogenides with modified Becke–Johnson density potential

In this work, we present first principles calculations based on a full potential linear augmented plane-wave method (FP-LAPW) to calculate structural and electronic properties of CdX and ZnX (X = S, Se, Te) based II–VI chalcogenides. First principles calculations using the local density approximation (LDA) and the related generalized gradient approximation (GGA) lead to a severe underestimate of the band gap. The proposed model uses various exchange–correlation potentials (LSDA, GGA and MBJLDA) to determine band gaps and structural properties of semiconductors. We show that using modified Becke–Johnson (MBJLDA) density potential leads to a better agreement with experimental data for band gaps of Cd and Zn based semiconductors.     

Implementation of LDA+DMFT with the pseudo-potential-plane-wave method

We propose an efficient implementation of combining dynamical mean field theory(DMFT) with electronic structural calculation based on the local density approximation(LDA).The pseudo-potential-plane-wave method is used in the LDA part,which enables it to be applied to large systems.The full loop self consistency of the charge density has been reached in our implementation,which allows us to compute the total energy related properties.The procedure of LDA+DMFT is introduced in detail with a complete flow chart.We have also applied our code to study the electronic structure of several typical strong correlated materials,including cerium,americium and NiO.Our results fit quite well with both the experimental data and previous studies.     

Ab initio study of structural, electronic and optical properties of Ca1−xSrxS compounds

Full-potential linearized augmented plane wave method (FP-LAPW) within density functional theory has been used to calculate structural, electronic and optical properties of Ca_1−xSr_xS, an alkali earth chalcogenide, with varying compositional parameter x in the range 0<x<1. Whereas the structural properties are discussed in terms of charge transfer between the two cations, calculated electronic band structure and density of states have been analyzed in terms of contribution from the S p, Ca d and Sr d states. Furthermore, we have calculated some optical properties such as real and imaginary parts of dielectric constant, ε(ω), and the calculated results have been discussed in comparison with the existing experimental data and other theoretical calculations.     

基于主题的自适应、在线网络热点发现方法及新闻推荐系统

 本文提出了一种基于改进HotRank算法的站点排序及种子URL选择方法,建立了在线主题发现系统信息采集自适应增量更新模型;结合LDA模型和仿射传播聚类算法(AP),提出了一种网络主题发现和热点新闻推荐方法,并在海天园知识服务平台热点新闻推荐系统中得到了应用.     

3-取代-3-苯基-1-茚酮类化合物的合成

以3-苯基-1-茚酮（1）为原料,THF为溶剂,二异丙基氨基锂（LDA）为碱,与碘代烷（RI）经亲核取代反应合成了4个3-取代-3-苯基-1-茚酮类化合物（3a~3d),其中3b~3d为新化合物,其结构经¹H NMR, ¹³C NMR 和EI-GC-MS表征。在最佳反应条件［1 1.0 eq., LDA/RI=2.1/1.5（当量比）,于0 ℃~rt反应5 h］下,3a~3d收率为70％～80％。     

激光诱导击穿光谱技术用于抹茶和绿茶粉的快速鉴别

研究利用激光诱导击穿光谱技术结合化学计量学方法快速鉴别抹茶和绿茶粉的可行性。抹茶与绿茶粉的主要区别在于茶树品种、栽培管理、生长时间和加工工艺。通过采集不同厂家生产的抹茶和不同杀青方式制成的绿茶粉在230～880nm的激光诱导击穿光谱并进行归一化预处理后,选用主成分分析(PCA),依据X-variables loadings获取用于鉴别抹茶和绿茶粉的特征波长,并基于特征波长建立线性判别式分析(LDA)模型。结果表明：基于特征波长建立的LDA模型能快速鉴别抹茶和绿茶粉,4个特征波长分别属于C(Ⅰ) 247.94 nm,Mg(Ⅱ) 279.60 nm,Ca(Ⅱ) 393.45 nm和Fe(Ⅱ) 766.68 nm;建模集和预测集的判别正确率均达到100%。采用激光诱导击穿光谱技术可以准确鉴别不同厂家生产的抹茶和不同杀青方式制成的绿茶粉。