检索结果-内蒙古大学图书馆

IEEE International Conference on Intelligence and Security Informatics, ISI

作者： Jie Wang Jun Zhang Laboratory for High Performance Scientific Computing and Computer SimulationDepartment of Compute Sciences University of Kentucky Lexington KY USA

Maintaining data mining accuracy on distorted datasets is an important issue in privacy preserving data mining. Using matrix approximation, we propose several efficient and flexible techniques to address this issue, and utilize unique characteristics of matrix factorization to maintain data pattern. We use the support vector machine classification to compare accuracy maintenance after data distortion by different methods. With better performance than some classical data perturbation approaches, nonnegative matrix factorization and singular value decomposition are considered to be promising techniques for privacy preserving data mining. Experimental results demonstrate that mining accuracy on the distorted data used these methods is almost as good as that on the original data, with added property of privacy preservation. It indicates that the matrix factorization-based data distortion schemes perturb only confidential attributes to meet privacy requirements while preserving general data pattern for knowledge extraction.

关键词： Data privacy Data mining Matrix decomposition Support vector machine classification Classification algorithms Support vector machines Additive noise Laboratories scientific computing computer simulation

来源：评论

学校读者我要写书评

暂无评论

Numerical approaches for dipole modeling in finite element method based source analysis

引用

International Congress Series 2007年 1300卷 189-192页

作者： Wolters, C.H. Köstler, H. Möller, C. Härdtlein, J. Anwander, A. Institute for Biomagnetism and Biosignalanalysis Münster University Hospital Münster Germany Scientific Computing and Imaging Institute University of Utah Salt Lake City United States Department of Computer Science 10 University Erlangen-Nuremberg Erlangen Germany Max Planck Institute for Human Cognitive and Brain Sciences Leipzig Germany

In EEG/MEG source analysis, a mathematical dipole is widely used as the "atomic" structure of the primary current distribution. When using realistic finite element models for the forward problem, the current dipole introduces a singularity on the right-hand side of the governing differential equation that has to be treated specifically. We evaluated and compared three different numerical approaches, a subtraction method, a direct approach using partial integration and a direct approach using the principle of Saint Venant. Evaluation and comparison were carried out in a four-layer sphere model using quasi-analytical formulas. © 2007 Elsevier B.V. All rights reserved.

关键词： Dipole Direct potential approach EEG Finite Element method MEG Singularity treatment Source analysis Subtraction method

来源：评论

学校读者我要写书评

暂无评论

NNMF-based factorization techniques for high-accuracy privacy protection on non-negative-valued datasets

NNMF-based factorization techniques for high-accuracy privac...

引用

作者： Wang, Jie Zhong, Weijun Zhang, Jun Department of Management Sciences and Engineering Southeast University Nanjing 210096 China Laboratory for High Performance Scientific Computing and Computer Simulation Department of Computer Sciences University of Kentucky Lexington KY 4056 United States

ISBN: (纸本)0769527027

The challenge in preserving data privacy is how to protect attribute values without jeopardizing the similarity between data objects under analysis. In this paper, we further our previous work on applying matrix techniques to protect privacy and present a novel algebraic technique based on iterative methods for non-negative-valued data distortion. As an unsupervised learning method for uncovering latent features in high-dimensional data, a low rank nonnegative matrix factorization (NNMF) is used to preserve natural data non-negativity and avoid subtractive basis vector and encoding interactions present in techniques such as principal component analysis. It is the first in privacy preserving data mining in our paper that combining non-negative matrix decomposition with distortion processing. Two iterative methods to solve bound-constrained optimization problem in NMF are compared by experiments on Wisconsin Breast Cancer Dataset. The overall performance of NMF on distortion level and data utility is compared to our previously-proposed SVD-based distortion strategies and other existing popular data perturbation methods. Data utility is examined by cross validation of a binary classification using the support vector machine. Our experimental results on data mining benchmark datasets indicate that, in comparison with standard data distortion techniques, the proposed NMF-based method are very efficient in balancing data privacy and data utility, and it affords a feasible solution with a good promise on high-accuracy privacy preserving data mining. © 2006 IEEE.

关键词： Non-negative matrix factorization

来源：评论

学校读者我要写书评

暂无评论

Adaptive Surface Reconstruction from Non Uniform Point Sets

Adaptive Surface Reconstruction from Non Uniform Point Sets

引用

International Conference on computer Engineering and Systems

作者： M. S. Abdel-wahab A. S. Hussein M. S. Gaber Scientific Computing Department Faculty of Computer and Information Sciences Ain Shams University Cairo Egypt

ISBN: (纸本)1424402719;1424402727

In this paper, a proposed algorithm for surface reconstruction from uniform or non-uniform point sets is introduced. The points are typically acquired with multiple range scans of any 3D object. The proposed algorithm follows the advancing front paradigm to build the reconstructed surface employing a variable radius moving ball that expands and shrinks continuously based on the sampling density. Starting with a user-specified initial radius, this initial ball may touch three points without containing any other point forming a seed triangle. For any edge, another point is found to form a ball with minimum radius generating another triangle. The process continues until generating all possible edges. The algorithm is theoretically proved under certain sampling criteria on the input data set. The proposed algorithm was applied on different datasets and compared favorably with the most eminent techniques. The key issues for comparisons were the reconstructed surface quality, the memory usage and the execution time. The present algorithm bested others in treating non uniform samples, samples with sharp edges and samples with small holes

关键词： Surface reconstruction Sampling methods Image reconstruction Surface treatment Shape Data visualization Surface texture Mesh generation Interpolation scientific computing

来源：评论

学校读者我要写书评

暂无评论

NNMF-Based Factorization Techniques for High-Accuracy Privacy Protection on Non-negative-valued Datasets

NNMF-Based Factorization Techniques for High-Accuracy Privac...

引用

IEEE International Conference on Data Mining Workshops (ICDM Workshops)

作者： Jie Wang Weijun Zhong Jun Zhang Department of Management Sciences and Engineering South-East University Nanjing China Laboratory for High Performance Scientific Computing and Computer Simulation Department of Computer Sciences University of Kentucky USA

关键词： Data privacy Protection Iterative methods Matrix decomposition Data analysis Unsupervised learning Encoding Principal component analysis Optimization methods Breast cancer

来源：评论

学校读者我要写书评

暂无评论

Improved string matching algorithm

引用

Jisuanji Gongcheng/computer Engineering 2006年第2期32卷 62-64页

作者： Wang, Cheng Liu, Jingang Joint Faculty of Computer Scientific Research Capital Normal University Beijing 100037 China Institute of Computing Technology Chinese Academy of Sciences Beijing 100080 China

String matching algorithm is an important method in intrusion detection. An improved algorithm is proposed for string matching based on the discussions of several common string matching algorithms, for example, BF, KMP, BM, Sunday, etc. This algorithm can jump over more characters to do next matching when this matching is lost. At the same time, the matching times decreases sharply. In this case, it improves the matching efficiency. In the end, the analysis of this algorithm's performance and the comparisons of the testing results of several matching algorithms are given.

关键词： Algorithm Pattern matching String searching

来源：评论

学校读者我要写书评

暂无评论

Parallel PDE-based simulations using the common component architecture

引用

Lecture Notes in Computational Science and Engineering 2006年 51卷 327-381页

作者： McInnes, Lois Curfman Allan, Benjamin A. Armstrong, Robert Benson, Steven J. Bernholdt, David E. Dahlgren, Tamara L. Diachin, Lori Freitag Krishnan, Manojkumar Kohl, James A. Walter Larson, J. Lefantzi, Sophia Nieplocha, Jarek Norris, Boyana Parker, Steven G. Ray, Jaideep Zhou, Shujia Mathematics and Computer Science Division Argonne National Laboratory Argonne IL United States Livermore CA United States Computer Science and Mathematics Division Oak Ridge National Laboratory Oak Ridge TN United States Center for Applied Scientific Computing Lawrence Livermore National Laboratory Livermore CA United States Computational Sciences and Mathematics Pacific Northwest National Laboratory Richland WA United States Reacting Flow Research SNL Livermore CA United States Scientific Computing and Imaging Institute University of Utah Salt Lake City UT United States Advanced Software R and D SNL Livermore CA United States Northrop Grumman Corporation Information Technology Sector Chantilly VA United States

ISBN: (纸本)9783540290766

The complexity of parallel PDE-based simulations continues to increase as multimodel, multiphysics, and multi-institutional projects become widespread. A goal of component - based software engineering in such large-scale simulations is to help manage this complexity by enabling better interoperability among various codes that have been independently developed by different groups. The Common Component Architecture (CCA) Forum is defining a component architecture specification to address the challenges of high-performance scientific computing. In addition, several execution frameworks, supporting infrastructure, and general-purpose components are being developed. Furthermore, this group is collaborating with others in the high-performance computing community to design suites of domain-specific component interface specifications and underlying implementations. © Springer-Verlag Berlin Heidelberg 2006.

关键词： Multiphysics

来源：评论

学校读者我要写书评

暂无评论

A hybrid system for three-dimensional objects reconstruction from point-clouds based on ball pivoting algorithm and radial basis functions

引用

WSEAS Transactions on computers 2005年第7期4卷 796-804页

作者： Hussein, Ashraf Scientific Computing Department Faculty of Computer and Information Sciences Ain Shams University Abbassia 11566 Cairo Egypt

In this paper, a new hybrid system for 3D objects reconstruction is presented. This system is based on the well-known Ball Pivoting Algorithm (BPA) of Bernardini et al. combined with the Radial Basis Function (RBF) model. The BPA can reconstruct surfaces from very large data sets in a little execution time, using small amount of memory, and exhibits linear complexity but it produces holes in case of low sampling density besides it is not capable to fill mis-registration domains. So, the BPA is coupled with the RBF in an integrated system to gain the edge of the RBF as a powerful interpolator for generating surface vertices within severe holes and/or mis-registration domains. Performance and accuracy of the system components are investigated through reconstructing benchmark objects. The results obtained compare favorably with other eminent algorithms. Finally, the proposed system exhibited high robustness in reconstructing objects from real large clouds with non-uniform sampling and mis-registrations.

关键词： Image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Some properties on multivariate filter banks with a matrix factorization

引用

Progress in Natural Science:Materials International 2005年第2期15卷 20-30页

作者： LIAN Qiaofang1, XIAO Hongying2* and CHEN Qiuhui3(1. Department of Mathematics, Beijing Jiaotong University, Beijing 100044, China 2. Institute of Mathematics, Academy of Mathematics and System sciences, Chinese Academy of sciences, Beijing 100080, China 3. Department of scientific computing and computer Applications, Zhongshan University, Guangzhou 510275, China) Department of Mathematics Beijing Jiaotong University Beijing 100044 China. Institute of Mathematics Academy of Mathematics and System Sciences Chinese Academy of Sciences Beijing 100080 China. Department of Scientific Computing and Computer Applications Zhongshan University Guangzhou 510275 China

A class of multivariate filter banks with a matrix factorization has been developed by Chen et al. The main purpose of this paper is to discuss some further properties of this kind of filter bank, such as the completeness, accuracy of the corresponding scaling functions and the relationships with Daubechies' wavelets and the multi-wavelets provided by Chui and Lian. Moreover, some examples are given to show that this kind of filter bank has a higher accuracy for the scaling functions.

关键词： filter banks completeness accuracy matrix factorization

来源：评论

学校读者我要写书评

暂无评论

Tree wavelet approximations with applications

引用

Science China Mathematics 2005年第5期48卷 680-702页

作者： XU Yuesheng & ZOU Qingsong Department of Mathematics, Syracuse University, Syracuse, NY 13244, USA Institute of Mathematics, Academy of Mathematics and System sciences, Chinese Academy of sciences, Beijing 100080, China Department of scientific computing and computer Science, Zhongshan University, Guangzhou 510275, China 1. Department of Mathematics Syracuse University 13244 Syracuse NY USA 2. Institute of Mathematics Academy of Mathematics and System Sciences Chinese Academy of Sciences 100080 Beijing China 3. Department of Scientific Computing and Computer Science Zhongshan University 510275 Guangzhou China

We construct a tree wavelet approximation by using a constructive greedy scheme (CGS). We define a function class which contains the functions whose piecewise polynomial approximations generated by the CGS have a prescribed global convergence rate and establish embedding properties of this class. We provide sufficient conditions on a tree index set and on bi-orthogonal wavelet bases which ensure optimal order of convergence for the wavelet approximations encoded on the tree index set using the bi-orthogonal wavelet bases. We then show that if we use the tree index set associated with the partition generated by the CGS to encode a wavelet approximation, it gives optimal order of convergence.

关键词： greedy algorithms tree wavelet approximations Besov spaces.

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：