检索结果-内蒙古大学图书馆

ICML 2006: 23rd International Conference on Machine Learning

作者： Hertz, Tomer Hillel, Aharon Bar Weinshall, Daphna School of Computer Science and Engineering Center for Neural Computation Hebrew University of Jerusalem Jerusalem 91904 Israel

ISBN: (纸本)1595933832

When given a small sample, we show that classification with SVM can be considerably enhanced by using a kernel function learned from the training data prior to discrimination. This kernel is also shown to enhance retrieval based on data similarity. Specifically, we describe KernelBoost - a boosting algorithm which computes a kernel function as a combination of 'weak' space partitions. The kernel learning method naturally incorporates domain knowledge in the form of unlabeled data (i.e. in a semi-supervised or transductive settings), and also in the form of labeled samples from relevant related problems (i.e. in a learning-to-learn scenario). The latter goal is accomplished by learning a single kernel function for all classes. We show comparative evaluations of our method on datasets from the UCI repository. We demonstrate performance enhancement on two challenging tasks: digit classification with kernel SVM, and facial image retrieval based on image similarity as measured by the learnt kernel.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

Application-adaptive Resource Scheduling in a computational Grid

引用

Journal of Zhejiang University-science B(Biomedicine & Biotechnology) 2006年第10期7卷 1634-1641页

作者： Cui-Ju Luan Guang-Hua Song Yao Zheng School of Computer Science and Center for Engineering and Scientific Computation Zhejiang University Hangzhou 310027 China

Selecting appropriate resources for running a job efficiently is one of the common objectives in a computational grid. Resource scheduling should consider the specific characteristics of the application, and decide the metrics to be used accordingly. This paper presents a distributed resource scheduling framework mainly consisting of a job scheduler and a local scheduler. In order to meet the requirements of different applications, we adopt HGSA, a Heuristic-based Greedy Scheduling Algorithm, to schedule jobs in the grid, where the heuristic knowledge is the metric weights of the computing resources and the metric workload impact factors. The metric weight is used to control the effect of the metric on the application. For different applications, only metric weights and the metric workload impact factors need to be changed, while the scheduling algorithm remains the same. Experimental results are presented to demonstrate the adaptability of the HGSA.

关键词： Grid Resource scheduling Heuristic knowledge Greedy scheduling algorithm

来源：评论

学校读者我要写书评

暂无评论

Portal Access to scientific computation and Customized Collaborative Visualization on Campus Grid

Portal Access to Scientific Computation and Customized Colla...

引用

IEEE Asia-Pacific Conference on Services Computing (APSCC)

作者： Yingjie Xia Guanghua Song Yao Zheng College of Computer Science and Center for Engineering and Scientific Computation University of Zhejiang Hangzhou China

In order to integrate the distributed computing resources in Zhejiang University (ZJU), a campus grid is constructed by using the hierarchical combination of middleware, Globus Toolkit and Sun Grid Engine. As one realization of the computational grid, the foundation of the campus grid provides a large amount of scientific computation cycles. Furthermore, our grid system adopts two different approaches, Java 3D and ParaView, to manipulate a collaborative visualization of the computed data respectively. Therefore the users can choose the most appropriate method according to their requirements. Three cases are tested on the campus grid via portal access. The results demonstrate that the ZJU campus grid provides powerful computation cycles and an efficient collaborative visualization service

关键词： Portals Grid computing Collaboration Data visualization Distributed computing Middleware Sun Engines Java Testing

来源：评论

学校读者我要写书评

暂无评论

Proxy-Based Parallel Visualization in a Grid Environment with PC Clusters

Proxy-Based Parallel Visualization in a Grid Environment wit...

引用

International Multi-Symposiums on computer and computational sciences (IMSCCS)

作者： Zhengge Huang Guanghua Song Yao Zheng College of Computer Science and Center for Engineering and Scientific Computation University of Zhejiang Hangzhou Zhejiang China

This paper describes our efforts to provide a grid-based parallel visualization environment to visualize massive dataset in parallel. The visualization environment is implemented as a visualization service on the grid. This paper focuses on deploying a proxy process on the master node of each cluster in the grid, to ensure that Globus jobs can be scheduled on internal nodes of clusters that have only local IP addresses. We have conducted an experiment to visualize in parallel a dataset of computational domains in a grid environment with PC clusters

关键词： Data visualization Concurrent computing Grid computing Resource management Software packages Educational institutions computer science Data engineering Processor scheduling Network servers

来源：评论

学校读者我要写书评

暂无评论

A Virtual Plant Ecosystem Featuring Parallel Computing and Distributed Visualization

A Virtual Plant Ecosystem Featuring Parallel Computing and D...

引用

International Multi-Symposiums on computer and computational sciences (IMSCCS)

作者： Zhengge Huang Yao Zheng Lijun Xie Tingjun Yang College of Computer Science and Center for Engineering and Scientific Computation University of Zhejiang Hangzhou Zhejiang China

A parallel plant ecosystem simulation, running on a computer cluster with commodity graphic cards, is performed to simulate and visualize large groups of plants. With a scalable architecture, the system can simulate very large and complex plant ecosystem in much shorter time than the traditional. The extreme complexity is first simplified by using multilevel models, and then dividing into multiple parts and simulating parallel. The entire simulation process is visualized in an immersive mode approximately in real time by rendering the results locally and concatenated to a large display wall

关键词： Ecosystems Parallel processing Visualization computational modeling computer simulation Concurrent computing computer graphics computer architecture Concatenated codes Displays

来源：评论

学校读者我要写书评

暂无评论

A Development Framework for Collaborative Applications Based on the Access Grid

A Development Framework for Collaborative Applications Based...

引用

International Multi-Symposiums on computer and computational sciences (IMSCCS)

作者： Yingjie Xia Yao Zheng Xuqing Zhu Chuihao Kong College of Computer Science and Center for Engineering and Scientific Computation University of Zhejiang Hangzhou Zhejiang China

Developing collaborative applications presents a challenge in solving the communication and synchronization problems. In this paper, we introduce a framework created to assist the developers. The framework takes the advantage of the access grid (AG) components, venue server, venue client and event channel, aiming to build a collaborative environment. In addition, the framework contains two user interface components, a wizard providing an interface to import new applications, and a main frame listing all the applications developed upon the framework. For a testing task, we implemented some collaborative applications using this framework. And the examples demonstrate that the framework enables the users to develop collaborative applications conveniently and efficiently

关键词： Collaboration Application software Collaborative work Collaborative software computer interfaces Resource management Educational institutions computer science Grid computing User interfaces

来源：评论

学校读者我要写书评

暂无评论

A General Framework for Parallel Planar Mesh Generation

A General Framework for Parallel Planar Mesh Generation

引用

International Multi-Symposiums on computer and computational sciences (IMSCCS)

作者： Ligang Chen Yi Liang Jianjun Chen Yao Zheng College of Computer Science and Center for Engineering and Scientific Computation University of Zhejiang Hangzhou Zhejiang China

In this paper, we present a general framework for 2D parallel mesh generation. A prepartitioner for domain decomposition is integrated into the framework, which strives to make the resulting subdomains well-shaped and thus guarantees the high quality of resulting meshes. Moreover, the time-consuming prepartitioning stage is parallelized in two ways with various grained levels, and the finer grained one turns out preferable. The subdomain connections are stored as a subdomain graph (SDG), which helps prevent mapping disconnected subdomains into a single processor whenever static or dynamic load balancing strategies are adopted. Well partitioned meshes could be generated simultaneously with parallel mesh generation, and hence the cost of mesh repartitioning could be eliminated or reduced. The SDG construction is simplified and independent of domain decomposition by introducing the concept of the characteristic polygon set (CPS). Shared nodes between neighboring subdomains are duplicated, that ensures subdomain meshing be completed with little or without communications, and full code-reuse of serial meshing algorithms be achieved in this framework

关键词： Mesh generation Geometry Concurrent computing Load management Costs computational fluid dynamics Partitioning algorithms Educational institutions computer science Numerical simulation

来源：评论

学校读者我要写书评

暂无评论

Embedding heterogeneous data using statistical models

Embedding heterogeneous data using statistical models

引用

21st National Conference on Artificial Intelligence and the 18th Innovative Applications of Artificial Intelligence Conference, AAAI-06/IAAI-06

作者： Globerson, Amir Chechik, Gal Pereira, Fernando Tishby, Naftali Computer Science and Artificial Intelligence Laboratory MIT Cambridge MA 02139 Computer Science Department Stanford University Stanford CA 94305 United States Department of Computer and Information Science University of Pennsylvania Philadelphia PA 19104 United States School of Computer Science and Engineering Interdisciplinary Center for Neural Computation Hebrew University Jerusalem 91904 Israel

ISBN: (纸本)1577352815

Embedding algorithms are a method for revealing low dimensional structure in complex data. Most embedding algorithms are designed to handle objects of a single type for which pairwise distances are specified. Here we describe a method for embedding objects of different types (such as authors and terms) into a single common Euclidean space based on their co-occurrence statistics. The joint distributions of the heterogenous objects are modeled as exponentials of squared Euclidean distances in a low-dimensional embedding space. This construction links the problem to convex optimization over positive semidefinite matrices. We quantify the performance of our method on two text datasets, and show that it consistently and significantly outperforms standard methods of statistical correspondence modeling, such as multidimensional scaling and correspondence analysis. Copyright © 2006, American Association for Artificial Intelligence (***). All rights reserved.

关键词： Data reduction

来源：评论

学校读者我要写书评

暂无评论

Ontology Construction for scientific Visualization

Ontology Construction for Scientific Visualization

引用

International Multi-Symposiums on computer and computational sciences (IMSCCS)

作者： Lijun Xie Yao Zheng Bin Shen College of Computer Science Center for Engineering and Scientific Computation Zhejiang University Hangzhou Zhejiang P. R. China

Ontology for scientific visualization is constructed in our laboratory. Its framework, construction method and design trade-off decision are presented in this paper. The ontology is composed of four main pillars: dataset, filter, device and pipeline, and they are then further decomposed into many sub-components. Using this concept of ontology can help scientists and engineers in various aspects such as communicating and sharing, quick learning, knowledge reusing, etc. As an application, we developed a semantic grid service for scientific visualization based on this ontology

关键词： Ontologies Data visualization Vocabulary Humans Pipelines Collaboration Grid computing Terminology Educational institutions computer science

来源：评论

学校读者我要写书评

暂无评论

Approximate dynamic programming based on expansive projections

Approximate dynamic programming based on expansive projectio...

引用

45th IEEE Conference on Decision and Control 2006, CDC

作者： Arruda, Edilson F. Do Val, João B. R. Center for Systems and Control National Laboratory for Scientific Computation - LNCC Av. Getulio Vargas 333 Petrópolis RJ 25651-075 Brazil School of Electrical and Computer Engineering State University of Campinas - UNICAMP Caixa Postal 6101 Campinas SP 13083-852 Brazil

ISBN: (纸本)1424401712

We present a general method to obtain convergent approximate value iteration algorithms with function approximation. The result is applicable to any arbitrary approximation architecture and generalizes existing results in the literature derived for particular approximation schemes. Additionally, we show how to obtain a convergent approximate mapping whose fixed point is the projection in the approximation space of a fixed point of the exact dynamic programming mapping with regards to a suitable subset norm. This result relies on evaluating the difference between successive iterates in the selected subset norm, which provides convergent procedures for any arbitrary approximation architecture. ©2006 IEEE.

关键词： Dynamic programming

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：