基于密度的kNN文本分类器训练样本裁剪方法

随着WWW的迅猛发展，文本分类成为处理和组织大量文档数据的关键技术。kNN方法作为一种简单、有效、非参数的分类方法，在文本分类中得到广泛的应用。但是这种方法计算量大，而且训练样本的分布不均匀会造成分类准确率的下降。针对kNN方法存在的这两个问题，提出了一种基于密度的kNN分类器训练样本裁剪方法，这种方法不仅降低了kNN方法的计算量，而且使训练样本的分布密度趋于均匀，减少了边界点处测试样本的误判。实验结果显示，这种方法具有很好的性能。     

基于kNN的快速WEB文档分类

根据测试文档在各个样本类中的分布情况，发现了基于kNN分类的两个有助于减少大量计算的重要性质，在此基础上提出了两个重要算法：排类算法和归类算法，从而构建了一个基于kNN的快速Web文档分类方法．理论与实验表明，这种方法可以在不影响原有准确率的条件下，可提高文档的分类速度．     

具有(k,r)-正交的(g,f)-因子分解的子图

研究了图的正交因子分解，通过构造函数p(x)和q(x)，证明了(mg k，mf-k)-图具有子图，该图有(g，f)-因子分解与kr-星(k，r)-正交，从而推广了原晋江教授的关于(mg m-1，mf-m 1)-图，存在(g，f)-因子分解与星(m，r)-正交的结论．     

网络处理器体系结构分析

该文旨在分析网络处理器能够同时满足高性能和灵活性要求的体系结构。而传统的网络设备单纯采用专用芯片或者基于RISC的通用处理器(GPPs),很难兼顾这两者要求。该文根据网络处理器的处理空间,将其映射为5个逻辑模块,这些模块由网络处理器中各个功能部件实现。然后分析了网络处理器的SMP和Pipeline两种并行结构,并进一步分析了隐藏延迟等实现加速的技术。最后分析了网络应用发展变化对网络处理器体系结构设计的挑战,并提出了解决办法。     

Improved analysis of sinigrin in Ethiopian mustard

An ion‐pair reversed‐phase high‐performance liquid chromatography (HPLC) method was developed for the analysis of sinigrin in Ethiopian mustard (Brassica carinata A. Braun) seed and seed fractions. Separation was compared on several RP‐HPLC columns (Inertsil^® ODS‐4 C₁₈ and ZORBAX^® Eclipse XDB‐C₁₈) with an isocratic eluent containing 100% aqueous (aq.) tetramethylammonium bromide (10 mm , pH 5.0). Sinigrin retention was affected by HPLC variables including the type of ion‐pair reagent, buffer strength and pH, acetonitrile concentration, column temperature and eluent flow rate. Partial validation demonstrated this optimised chromatographic condition to be linear, accurate and precise. Multistage extraction using 70% (v/v) aq. methanol was more efficient than 50% (v/v) aq. acetonitrile. In addition, the matrix effect and recovery rate as well as processing efficiency of the analytical protocol were determined. This method is suitable for high throughput analysis of sinigrin in Ethiopian mustard seed and seed fractions.     

A scheme for LR(k) parsing with error recovery part II: error recovery

The paper is the second in a series of three papers devoted to a detailed study of LR(k) parsing with error recovery and correction. Error recovery in LR(k) parsing of a context-free grammar is formalized by extending an LR(k) parser of the grammar such that it accepts all strings over the terminal vocabulary. The parse produced by this extension for a terminal string is a right parse if the string is in the language. In the case of a string not in the language the parse produced by the extension contains so-called error productions which represent the error recovery actions performed by the extension. The treatment is based on the formalization of LR(k) parsing presented in the first paper in the series and it covers practically all error recovery methods designed for LR(k) parsing.     

Viewers’ optimization of preferred viewing distance by spatial resolution of TV display

We have shown in a previous paper that preferred viewing distance (PVD)—a viewer’s favored distance from which to watch TV in the home viewing environment—is shorter when watching a high spatial resolution 4k TV than on a high-definition TV (HDTV). PVD depends on (1) the spatial resolution of the viewer’s object of interest in the relevant scene, (2) whether viewers have sufficient time to optimize their PVD, as with still images, and (3) whether they are appropriately encouraged to pay attention to the object at hand on the screen, by, for example, narrative instructions. In this paper, we demonstrate that an analogous result is obtained in a comparison between high spatial resolution 8k TV and HDTV: The PVD of viewers for 8k TV is shorter than that for HDTV. This result suggests new possibilities for high spatial resolution TV, such as novel applications and TV programs that are especially developed for 8k TV systems.     

Stabilizing effect of the cyclodextrins additive to spray-dried particles of curcumin/polyvinylpyrrolidone on the supersaturated state of curcumin

Although curcumin is considered to have various therapeutic effects, its use as a functional food or supplement is restricted owing to its low water solubility and bioavailability. To increase the solubility of curcumin in water, the use of polyvinylpyrrolidone (PVP) and vinylpyrrolidone-vinyl acetate copolymers with a pyrrolidone skeleton was noted to be promising. In particular, the bi-component formulations of curcumin/PVP prepared through spray drying exhibited an amorphous state in powder X-ray diffraction observations and temporally increased the apparent solubility of curcumin to over 5000 times that of untreated curcumin; nevertheless, after 24 h, the solubility decreased owing to the unstable supersaturated state of curcumin. The addition of α-cyclodextrin (α-CyD) in the bi-component curcumin/PVP formulation helped maintain the supersaturated state of curcumin, whereas the addition of β- and γ-CyD led to the collapse of the supersaturated state. The addition of α-CyD can likely help inhibit the nucleation and crystal growth of curcumin, through the interaction among the solubilized units of curcumin/PVP and α-CyD.     

Coastal trapped waves in the southern Caspian Sea

Previous studies reported coastal trapped waves (CTWs) in the Caspian Sea (CS). This study deals with the generation mechanisms, the temporal and spatial variability of CTWs in this area, and their transformations during propagation from the origin to the destination using recent measurements and high-resolution numerical simulations. CTWs are observed at all stations with periods of 2–6 days after northerly storms. The Absheron Peninsula, old Sefidrud delta, and Nur coasts were identified as the CTWs prone regions. The generation of CTWs in these locations was confirmed using numerical experiments. The propagation away from the generating location of CTWs was analyzed using a representative real wind storm. In the west part of the CS, the generation mechanism of CTWs is mainly similar to the barotropic Kelvin waves; in contrast, it is similar to the continental shelf waves in the southern shelves. The results can be used to study the contribution of generated CTWs to the transport of sediment and biological materials in all large lakes.     

Design of data feature-driven 1D/2D convolutional neural networks classifier for recycling black plastic wastes through laser spectroscopy

In this study, two types of convolutional neural network (CNN) classifiers are designed to handle the problem of classifying black plastic wastes. In particular, the black plastic wastes have the property of absorbing laser light coming from spectrometer. Therefore, the classification of black plastic wastes remains still a challenging problem compared to classifying other colored plastic wastes using existing spectroscopy (i.e., NIR). When it comes the classification problem of black plastic wastes, effective classification techniques by the laser spectroscopy of Fourier Transform-Infrared Radiation (FT-IR) with Attenuated Total Reflectance (ATR) and Raman to analyze the classification problem of black plastic wastes are introduced. Due to the strong ability of extracting spatial features and remarkable performance in image classification, 1D and 2D CNN through data features are designed as classifiers. The technique of chemical peak points selection is considered to reduce data redundancy. Furthermore, through the selection of data features based on the extracted 1D data with peak points is introduced. Experimental results demonstrate that 2DCNN classifier designed with the help of 2D data feature selection as well as 1DCNN classifier shows the best performance compared with other reported methods for classifying black plastic wastes.