检索结果-内蒙古大学图书馆

20th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2006

作者： Quinlan, Dan Schordan, Markus Vuduc, Richard Yi, Qing Center for Applied Scientific Computing Lawrence Livermore National Laboratory Livermore CA 94550 United States Institute of Computer Languages Vienna University of Technology Vienna Austria Dept. of Computer Science University of Texas at San Antonio San Antonio TX 78249 United States

ISBN: (纸本)1424400546

Although conventional compilers implement a wide range of optimization techniques, they frequently miss opportunities to optimize the use of abstractions, largely because they are not designed to recognize and use the relevant semantic information about such abstractions. In this position paper, we propose a set of annotations to help communicate high-level semantic information about abstractions to the compiler, thereby enabling the large body of traditional compiler optimizations to be applied to the use of those abstractions. Our annotations explicitly describe properties of abstractions that are needed to guarantee the applicability and profitability of a broad variety of such optimizations, including memoization, reordering, data layout transformations, and Mining and specialization. © 2006 IEEE.

关键词： Program compilers

来源：评论

学校读者我要写书评

暂无评论

Enabling scientific Workflow Reuse through Structured Composition of Dataflow and Control-Flow 22

Enabling Scientific Workflow Reuse through Structured Compos...

引用

22nd International Conference on Data Engineering Workshops, ICDEW 2006

作者： Bowers, Shawn Ludäscher, Bertram Ngu, Anne H.H. Critchlow, Terence UC Davis Genome Center University of California Davis United States Deptartment of Computer Science UC Davis Genome Center University of California Davis United States Department of Computer Science Texas State University San Marcos United States Center for Applied Scientific Computing Lawrence Livermore National Laboratory United States

ISBN: (纸本)0769525717

Data-centric scientific workflows are often modeled as dataflow process networks. The simplicity of the dataflow framework facilitates workflow design, analysis, and optimization. However, modeling "control-flow intensive" tasks using dataflow constructs often leads to overly complicated workflows that are hard to comprehend, reuse, and maintain. We describe a generic framework, based on scientific workflow templates and frames, for embedding control-flow intensive subtasks within dataflow process networks. This approach can seamlessly handle complex control-flow without sacrificing the benefits of dataflow. We illustrate our approach with a real-world scientific workflow from the astrophysics domain, requiring remote execution and file transfer in a semi-reliable environment. For such workflows, we also describe a 3-layered architecture based on frames and templates where the top-layer consists of an overall dataflow process network, the second layer consists of a tranducer template for modeling the desired control-flow behavior, and the bottom layer consists of frames inside the template that are specialized by embedding the desired component implementation. Our approach can enable scientific workflows that are more robust (faulttolerance strategies can be defined by control-flow driven transducer templates) and at the same time more reusable, since the embedding of frames and templates yields more structured and modular workflow designs. © 2006 IEEE.

关键词： Astrophysics

来源：评论

学校读者我要写书评

暂无评论

Modified V-BLAST Symbol Detection Under Channel Uncertainties for MIMO Systems

Modified V-BLAST Symbol Detection Under Channel Uncertaintie...

引用

Asilomar Conference on Signals, Systems & computers

作者： Hyun Jong Yang Joohwan Chun Scientific Computing Laboratory Department of Electrical Engineering and Computer Science Korea Advanced Institute of Science and Technology Daejeon South Korea

The performance of V-BLAST symbol detection can be seriously degraded if channel information is not perfect. We derive a new nulling matrix at the receiver solving the min-max problem which is robust to the statistical changes of channel error. Using the new nulling matrix, we propose a modified vertical Bell laboratories layered space-time(V-BLAST) detection algorithm to reduce unexpected effects of channel uncertainties. The proposed algorithm requires only one more scalar value, the maximal norm of a channel error matrix, than the conventional algorithm does, but the performance is better especially at the high SNR. We simulate the proposed algorithm under correlated MIMO channels with Gaussian channel coefficients.

关键词： Uncertainty MIMO Covariance matrix Gaussian noise Additive noise Upper bound Cost function Receiving antennas scientific computing Degradation

来源：评论

学校读者我要写书评

暂无评论

Large-scale electronic structure calculations of high-Z metals on the BlueGene/L platform

Large-scale electronic structure calculations of high-Z meta...

引用

2006 ACM/IEEE Conference on Supercomputing, SC'06

作者： Gygi, Francois Draeger, Erik W. Schulz, Martin De Supinski, Bronis R. Gunnels, John A. Austel, Vernon Sexton, James C. Franchetti, Franz Kral, Stefan Ueberhuber, Christoph W. Lorenz, Juergen Department of Applied Science University of California Davis Davis CA 95616 Center for Applied Scientific Computing Lawrence Livermore National Laboratory Livermore CA 94551 IBM Thomas J. Watson Research Center Yorktown Heights NY 10598 Department of Electrical and Computer Engineering Carnegie Mellon University Pittsburgh PA 15213 Institute of Analysis and Scientific Computing Vienna University of Technology Vienna Austria

ISBN: (纸本)0769527000

First-principles simulations of high-Z metallic systems using the Qbox code on the BlueGene/L supercomputer demonstrate unprecedented performance and scaling for a quantum simulation code. Specifically designed to take advantage of massively-parallel systems like BlueGene/L, Qbox demonstrates excellent parallel efficiency and peak performance. A sustained peak performance of 207.3 TFlop/s was measured on 65,536 nodes, corresponding to 56.5% of the theoretical full machine peak using all 128k CPUs. © 2006 IEEE.

关键词： Electronic structure

来源：评论

学校读者我要写书评

暂无评论

Combining Generative and Discriminative Methods for Pixel Classification with Multi-Conditional Learning

Combining Generative and Discriminative Methods for Pixel Cl...

引用

International Conference on Pattern Recognition

作者： B.M. Kelm C. Pal A. McCallum Interdisciplinary Center for Scientific Computing University of Heidelberg Germany Department of Computer Science University of Massachusetts Amherst USA

It is possible to broadly characterize two approaches to probabilistic modeling in terms of generative and discriminative methods. Provided with sufficient training data the discriminative approach is expected to yield superior accuracy as compared to the analogous generative model since no modeling power is expended on the marginal distribution of the features. Conversely, if the model is accurate the generative approach can perform better with less data. In general it is less vulnerable to overfitting and allows one to more easily specify meaningful priors on the model parameters. We investigate multi-conditional learning - a method combining the merits of both approaches. Through specifying a joint distribution over classes and features we derive a family of models with analogous parameters. Parameter estimates are found by optimizing an objective function consisting of a weighted combination of conditional log-likelihoods. Systematic experiments in the context of foreground/background pixel classification with the Microsoft-Berkeley segmentation database using mixtures of factor analyzers illustrate tradeoffs between classifier complexity, the amount of training data and generalization accuracy. We show experimentally that this approach can lead to models with better generalization performance than purely generative or discriminative approaches

关键词： Character generation Parameter estimation Training data Context modeling Linear discriminant analysis Logistics scientific computing computer science Power generation Power system modeling

来源：评论

学校读者我要写书评

暂无评论

Interactive display of isosurfaces with global illumination

Interactive display of isosurfaces with global illumination

引用

作者： Wyman, Chris Parker, Steven Shirley, Peter Hansen, Charles IEEE University of Iowa 14 MacLean Hall Iowa City IA 52242-1419 Scientific Computing and Imaging Institute University of Utah 50 S. Central Campus Dr. Salt Lake City UT 84112 School of Computing University of Utah 50 S. Central Campus Dr. Salt Lake City UT 84112 Department of Computer Science University of Iowa Institute University of Utah SCIRun Problem Solving Environment Uintah School of Computing University of Utah Department of Computer Science University of Utah

In many applications, volumetric data sets are examined by displaying isosurfaces, surfaces where the data, or some function of the data, takes on a given value, interactive applications typically use local lighting models to render such surfaces. This work introduces a method to precompute or lazily compute global illumination to improve interactive isosurface renderings. The precomputed illumination resides in a separate volume and includes direct light, shadows, and interreflections. Using this volume, interactive globally illuminated renderings of isosurfaces become feasible while still allowing dynamic manipulation of lighting, viewpoint and isovalue. © 2006 IEEE.

关键词： Surface properties

来源：评论

学校读者我要写书评

暂无评论

A Path-sequence Based Discrimination for Subtree Matching in Approximate XML Joins

A Path-sequence Based Discrimination for Subtree Matching in...

引用

International Conference on Data Engineering Workshops

作者： Wenxin Liang H. Yokota Department of Computer Science Tokyo Institute of Technology Tokyo Japan Department of Computer Science Global Scientific Information and Computing Center Tokyo Institute of Technology Tokyo Japan

In this paper, we discuss the one-to-multiple matching problem in leaf-clustering based approximate XML join algorithms and propose a path-sequence based discrimination method to solve this problem. In our method, each path sequence from the top node to the matched leaf in the base and target subtree is extracted, and the most similar target subtree for the base one is determined by the pathsequence based subtree similarity degree. We conduct experiments to evaluate our method by using both real bibliography and bioinformatics XML documents. The experimental results show that our method can effectively decrease the occunence rate of one-to-multiple matching for both bibliography and bioinformatics XML data, and hence improve the precision of the leaf-clustering based approximate XML join algorithms.

关键词： XML Bioinformatics Bibliographies Internet computer science Clustering algorithms Data engineering Conferences

来源：评论

学校读者我要写书评

暂无评论

Solutions of the 1D Coupled Nonlinear Schrodinger Equations by the CIP-BS Method

引用

Communications in Computational Physics 2006年第2期1卷 261-275页

作者： Takayuki Utsumi Takayuki Aoki James Koga Mitsuru Yamagiwa Department of Electronics and Computer Science Tokyo University of ScienceYamaguchi1-1 Daigaku-doriSanyou-OnodaYamaguchi 756-0884Japan Global Scientific Information and Computing Center Tokyo Institute of TechnologyO-okayamaMeguro-kuTokyo 152-8552Japan Japan Atomic Energy Agency 8-1 Umemidai Kizu-choSouraku-gunKyoto 619-0215Japan

In this paper,we present solutions for the one-dimensional coupled nonlinear Schrödinger(CNLS)equations by the Constrained Interpolation Profile-Basis Set(CIP-BS)*** method uses a simple polynomial basis set,by which physical quantities are approximated with their values and derivatives associated with grid *** operations on functions are carried out in the framework of differential ***,by introducing scalar products and requiring the residue to be orthogonal to the basis,the linear and nonlinear partial differential equations are reduced to ordinary differential equations for values and spatial *** method gives stable,less diffusive,and accurate results for the CNLS equations.

关键词： The CIP-BS method basis set differential algebra the Galerkin method the coupled nonlinear Schrödinger equation

来源：评论

学校读者我要写书评

暂无评论

On the optimal working set size in serial and parallel support vector machine learning with the decomposition algorithm

Conferences in Research and Practice in Information Technolo...

引用

Conferences in Research and Practice in Information Technology Series 2006年 61卷 121-128页

作者： Eitrich, Tatjana Lang, Bruno Central Institute for Applied Mathematics Research Centre Juelich Germany Applied Computer Science and Scientific Computing Group Department of Mathematics University of Wuppertal Germany

The support vector machine (SVM) is a wellestablished and accurate supervised learning method for the classification of data in various application fields. The statistical learning task - the so-called training - can be formulated as a quadratic optimization problem. During the last years the decomposition algorithm for solving this optimization problem became the most frequently used method for support vector machine learning and is the basis of many SVM implementations today. It is characterized by an internal parameter called working set size. Usually small working sets have been assigned. The increasing amount of data used for classification led to new parallel implementations of the decomposition method with efficient inner solvers. With these solvers larger working sets can be assigned. It was shown, that for parallel training with the decomposition algorithm large working sets achieve good speedup values. However, the choice of the optimal working set size for parallel training is not clear. In this paper, we show how the working set size influences the number of decomposition steps, the number of kernel function evaluations and the overall training time in serial and parallel computation. © 2006, Australian computer Society, Inc.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

A path-sequence based discrimination for subtree matching in approximate XML joins 22

A path-sequence based discrimination for subtree matching in...

引用

22nd International Conference on Data Engineering Workshops, ICDEW 2006

作者： Liang, Wenxin Yokota, Haruo Department of Computer Science 2-12-1 Oh-Okayama Meguro-ku Tokyo152-8552 Japan Global Scientific Information and Computing Center Tokyo Institute of Technology 2-12-1 Oh-Okayama Meguro-ku Tokyo152-8552 Japan

ISBN: (纸本)0769525717

In this paper, we discuss the one-To-multiple matching problem in leaf-clustering based approximate XML join algorithms and propose a path-sequence based discrimination method to solve this problem. In our method, each path sequence from the top node to the matched leaf in the base and target subtree is extracted, and the most similar target subtree for the base one is determined by the pathsequence based subtree similarity degree. We conduct experiments to evaluate our method by using both real bibliography and bioinformatics XML documents. The experimental results show that our method can effectively decrease the occunence rate of one-To-multiple matching for both bibliography and bioinformatics XML data, and hence improve the precision of the leaf-clustering based approximate XML join algorithms. © 2006 IEEE.

关键词： XML

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：