基于粗糙集理论的数据清洗模型

本文在基于粗糙集理论的基础上提出一个大数据系统的数据清洗模型,以利于信息不完备决策问题的实施。模型从给定的初始数据出发,在假定属性的重要性量度σ0的前提下,经过一系列的计算,比较属性实际重要性量度σCD（x)与σ0的大小,由此确定属性x的清洗,然后根据清洗后的数据的出相应的决策规则。并给出一个简单的例子来验证数据清洗模型。     

在属性级别上处理噪声数据的数据清洗算法

噪声数据的处理是数据清洗的一个重要环节.文章提出一个基于聚类分析的在属性级别上处理噪声数据的算法.该算法通过聚类分析确认噪声数据产生噪声的具体属性,同时统计噪声在属性上的分布规律,这些统计数据可以在下一步的工作中进一步提高数据质量.     

基于权重联系度粗集理论的增量式知识发现

在实际生活中,信息系统的增量数据会不断产生,如何充分利用以前计算的结果结合新产生的数据进行新的知识发现,这是有意义的。针对这样的问题,提出了基于权得联系度的粗集模型,它着重考虑了条件属性重要性存在差异来建立粗集模型,而条件属性重要性可以通过以前数据的知识发现计算出来,于是利用基于权重联系度的粗集模型在对新产生的数据进行知识发现时,利用了以前的数据信息。建立了基于权重联系度的粗集模型及其对应的属性和属性值约简理论,最后通过一个示例来演示增量式知识发现的方法。     

属性重要性评分方法的改进

研究信息系统的属性重要性评分方法,通过引入敏感系数构建神经网络模型,提出属性重要性评分算法,将信息系统的各条件属性和决策属性构造一个径向基函数(RBF)神经网络。经训练和学习后,综合考虑各属性间的关系,动态调整RBF网络的拓扑结构,评分各属性的重要性。以红籽西瓜性状数据作为样本数据和测试数据进行实例分析,验证该方法的有效性。     

基于决策分类的决策表分解方法研究*

针对当前基于属性重要性的决策表属性集分解方法存在的不足,提出了一种新型的基于决策分类的决策表属性集分解方法。分析了近似分类质量和属性重要性与决策分类之间的关系,利用粗糙集理论,从提高子决策表中决策分类正确性的角度出发考虑条件属性与决策属性之间的关系,提出了决策表分解的条件属性选择量度并对决策表实施属性集分解。     

基于云理论的数据属性约简

提出了一个基于云模型的数据属性约简新算法,该算法借助云模型将每个连续属性上的定量数据转化为定性概念,然后利用此定性概念计算每个条件属性对决策属性的重要性,最后以贪婪算法的思想求解约简属性,从而解决了当前约简算法对连续属性必须离散化的问题.同时对算法中可能遇到的冲突问题给出了初步的解决方案.实验表明,该算法在降低时间复杂度的同时能求出更好的属性集约简.     

使用Delphi经验三则

1．用Chmbo：IkZx构件实现数据控件DBCOmboBox的功能由于Delphi的DBComIX＞oc＞x控件没有Text属性，使其在某些场合编程时很不方便。而普通COmb0BoX控件虽有Text属性却无DataField属性，笔者通过将已m－boBox与Query控件的配合使用，使其成为既有Text属性又包含一个数据库表字段的下拉列表框。实例：某数据库系统要求用下拉列表框对一个名单表格的“NAME”字段的内容进行选择，以便实现系统按“姓名”进行动态查询。见下表：这里我们用ComboB0x配合Qllery控件来实现上述功能的下拉列表框，具体实现方法很简单，首先用Query在…     

Web数据清洗及其系统框架研究

数据清洗是提高数据质量的有效手段。分析了从Web上抽取的数据存在的质量问题或错误,针对错误类型,给出属性错误(包括不完整数据和异常数据)和重复与相似重复记录的描述,并提出相应的清洗方法;设计了一个数据清洗系统框架,该框架由数据预处理、数据清洗引擎和质量评估三大部分组成,可以针对不同的错误类型,完成不同的清洗任务。实验表明,该框架具有通用性和可扩展性。     

一种透明GIF图象的制作方法

GIF图象格式以它众多的精美图象而闻名,其89a版支持一种透明属性,在许多领域得到广泛应用,如Internet中Home Page的制作,最为常见的是87a格式的GIF图象.本文介绍了一种87aGIF图象转换为89a透明图的方法.1.GIF的文件结构GIF有五个主要分量以固定顺序出现:识别标志(GIF87a或GIF89a);逻辑屏幕描述块说明了用来生成的显示文件中的图象显示器分辨率,分别表示屏幕的宽度和高;全局色表(如果存在)构成一个24位RGB三联体的调色板;位图图象数据(标志为0x2c);数据结束符(标志为0x3b).对于89a格式全局色表的后边可以存在一个图象控制扩展块或其它扩展(扩展标志为0x21),透明属性由图象控制扩展描述的.     

一种不确定RFID数据流清洗策略

原始RFID数据流含有大量噪声且具有不确定性,必须在使用之前对其进行数据清洗,而清洗策略是清洗质量的保证。提出一种适合不确定RFID数据流的清洗策略。该清洗策略引入了最大熵原理,对待清洗的RFID元组的特征属性进行权重选择,并根据清洗节点的时间消耗以及误差进行清洗成本分析,决策出最佳的清洗方法。仿真实验结果表明,该清洗策略提高了不确定RFID数据流的清洗效率与精度。     

Data base design for computer-integrated structural engineering

The most likely architecture for an integrated engineering software system is a series of application programs operating off a shared central data base. The data base is the most critical part of the system. This article proposes a design for such a data base. The steps involved in data base, design are first reviewed, and several requirements for an effective structural engineering data base system are identified. A data base design based on a component-connection abstraction model and the object-oriented data model is then presented.     

An Interval Polynomial Interpolation Problem and Its Lagrange Solution

In many practical problems, we need to use interpolation: we know that the value of a quantity is uniquely determined by some other quantity x (i.e., y = f(x)), we have measured several pairs of values (xi, yi), and we want to predict y for a given x. We can only guarantee estimates for y if we have some a priori information about the function f(x). In particular, in some problems, we know that f(x) is a polynomial of known degree d (e.g., that it is linear, or that it is quadratic). For this polynomial interpolation, with interval uncertainty of the input data (xi, yi), we present several reasonable algorithms that compute, for a given x0, guaranteed bounds for f(x0).     

A model of the dynamic behavior of B-trees

Summary We present a practical and efficient model for the estimation of average performance measures of B-trees under dynamic conditions of insertions and deletions. Performance measures computed are average storage utilization, average path length, and average tree height. The model introduces a data structure, called a lineage tree, which permits a highly compact representation of B-trees while still retaining information needed to compute the above performance measures. The model then involves a Markov chain in which the states are lineages obtained from the lineage tree. Probabilities, based on the number of B-tree structures corresponding to each lineage, are derived for the transition from one lineage to another under certain dynamic conditions. Results are given for tree orders ranging from 5 up to 401, and for numbers of keys up to 140000. Computer requirements are shown to be small to moderate.     

A Study of a Convex Variational Diffusion Approach for Image Segmentation and Feature Extraction

We analyze a variational approach to image segmentation that is based on a strictly convex non-quadratic cost functional. The smoothness term combines a standard first-order measure for image regions with a total-variation based measure for signal transitions. Accordingly, the costs associated with discontinuities are given by the length of level lines and local image contrast. For real images, this provides a reasonable approximation of the variational model of Mumford and Shah that has been suggested as a generic approach to image segmentation.The global properties of the convex variational model are favorable to applications: Uniqueness of the solution, continuous dependence of the solution on both data and parameters, consistent and efficient numerical approximation of the solution with the FEM-method.Various global and local properties of the convex variational model are analyzed and illustrated with numerical examples. Apart from the favorable global properties, the approach is shown to provide a sound mathematical model of a useful locally adaptive smoothing process. A comparison is carried out with results of a region-growing technique related to the Mumford-Shah model.     

TMPR: A tree-structured modified problem reduction proof procedure and its extension to three-valued logic

This paper deals with automated deduction for classical and partial logics, especially for the three-valued logic L3, which has been introduced, for example, in the study of natural language semantics. Based on ideas from a Plaisted's Gentzen style system for classical two-valued logic, we present a new tree-structured proof procedure (TMPR) together with a new completeness proof using proof transformation techniques and some improvements including the generation and use of lemmata. TMPR extends SLD-resolution with a Prolog-style backward chaining to full first-order logic by a controlled use of case analysis. This is done without having to extend negative goals needed, for example, for model elimination. A classification of TMPR, model elimination and related calculi in a common tableau framework is given. Thereafter, we present our extension of the TMPR proof procedure to L3 and show its soundness and completeness. As a side result, a TMPR proof system for the four-valued logic L4 is given. Finally, the restriction of TMPR to L3-Horn clauses is considered, and, additionally, an idea for similarly extending model elimination and related systems to L3 (and L4) is illustrated.This work is supported by the KI-Verbund NRW, founded by the Ministry for Science and Research of North Rhine Westphalia and by the Deutsche Forschungs Gemeinschaft in the scope of the research topic Deduktion, and is an extended version of a talk held at the German-Japanese Workshop on Logic and Natural Language (23–25 October 1990, in Kyoto, Japan).     

Numerical Solution Of Inverse Problem For Elliptic Pdes

This work is concerned with computing the solution of the following inverse problem: Finding u and 𝜌on D such that: $$\nabla \cdot (\rho \nabla u) = 0,\quad \hbox{on}\ D;$$ $$u = g,\quad \hbox{on}\ \partial D;\qquad \rho u_n = f,\quad \hbox{on}\ \partial D;$$ $$\rho (x_0, y_0) = \rho_0,\quad \hbox{for a given point}\ (x_0, y_0) \in D$$ where f and g are two given continuous functions defined on the boundary of D , and D is a given bounded region of R 2 . The solution is found using a development of the direct variational method. The two unknown functions are represented by linear combinations of certain classes of functions and using multiobjective optimization to minimize the two objective functionals F and H , where $$F = \vint \vint_D \rho (x,y) \nabla u\cdot \nabla u\,\hbox{d}x\,\hbox{d}y\quad \hbox{and}\quad H = \vint_{\partial D} (\rho u_n - f)^2 \hbox{d}s$$ A computer program is written and implemented and tested for data formed by numerical simulation.     

Conservative difference schemes for diffusion problems with boundary and interface conditions

For 0≤x≤1, 0≤t≤T we consider the diffusion equation $$\gamma (x)u_t (x, t) - (B u)_x (x, t) = f(x, t)$$ with (alternatively)B u:=(a(x)u) _x +b(x)u ora(x)u _x +β(x)u. There are given initial valuesu(x,0), influx rates?(B u) (0,t) and (B u) (1,t) across the lateral boundaries and an influx rate (B u) (ζ?0,t)?(B u) (ζ+0,t) at an interface ζ∈(0, 1) where the elsewhere smooth functions γ,a, b, β are allowed to have jump discontinuities.a and γ are assumed to be positive. Interpretingu(x, t) as temperature and γ(x) u (x, t) as energy density we can easily express the total energy \(E(t) = \int\limits_0^1 {\gamma (x) u (x, t)} \) in terms of integrals of the given data. We describe and analyse explicit and implicit one-step difference schemes which possess a discrete quadrature analogue exactly matchingE(t) at the time grid points. These schemes also imitate the isotonic dependence of the solution on the data. Hence stability can be proved by Gerschgorin's method and, under appropriate smoothness assumptions, convergence is 0 ((Δx)²+Δt).     

3-D interpretation of optical flow by renormalization

This article studies 3-D interpretation of optical flow induced by a general camera motion relative to a surface of general shape. First, we describe, using the image sphere representation, an analytical procedure that yields an exact solution when the data are exact: we solve theepipolar equation written in terms of theessential parameters and thetwisted optical flow. Introducing a simple model of noise, we then show that the solution is statistically biased. In order to remove the statistical bias, we propose an algorithm calledrenormalization, which automatically adjusts to unknown image noise. A brief discussion is also given to thecritical surface that yields ambiguous 3-D interpretations and the use of theimage plane representation.     

不同蕴涵算子下三I方法的解

以R_M、R_Z、R₀三个蕴涵算子为基础,研究了11种形式的三I算法的解,并在此基础上给出了这11种解的同一形式：B*（y）=SUP{A*（x）∧φ（x,y）},其中φ：X×Y→[0,1]表示某一函数,而且φ（x,y）与E_Y的选取依赖蕴涵算子R_i的选取。     

The spectral density of a nonlinear damping model: multi-DOF case

In this work, the spectral density of the following multi-DOF nonlinear damping model is investigated: Mx¨+D₀x˙+γD(x, x˙)+Kx=σn(t) where γ>0 is a small parameter. A formula for the spectral density is established with O(γ2) accuracy based upon the Fokker-Planck technique and perturbation. One of the features of the multi-DOF oscillation system is that x and x˙ are generally correlated in stationary state. This is true even for linear systems. Necessary and sufficient conditions for uncorrelatedness are given for linear systems. Since the first-order statistics R_xx(0) and R_xy(0), where y=x˙, appear in the spectral density formula, it is desirable to have the explicit stationary probability density for the purpose of evaluating R_xx(0) and R_xy (0). However, in general, as in the single DOF case, an expression for the stationary density is not available. This note gives the explicit stationary density of an “energy”-type nonlinear damping model Mx¨+μ(E_D)Dx˙+Kx=σn(t) in which the “energy” E_D is defined as E_D=1/2(x ^TKDx+y^TMDy) where D>0 is assumed to commute with K and M. In the end, an energy-type nonlinear damping model is worked out completely as an illustration