检索结果-内蒙古大学图书馆

IEEE International Conference on Cluster Computing (CLUSTER)

作者： Subasi, Omer Di, Sheng Balaprakash, Prasanna Unsal, Osman Labarta, Jesus Cristal, Adrian Krishnamoorthy, Sriram Cappello, Franck Pacific Northwest Natl Lab Richland WA 99352 USA Barcelona Supercomp Ctr Barcelona Spain Argonne Natl Lab Lemont IL USA CSIC Spanish Natl Res Council IIIA Artificial Intelligence Res Inst Madrid Spain

ISBN: (纸本)9781538623268

Future high-performance computing (HPC) systems with ever-increasing resource capacity (such as compute cores, memory and storage) may significantly increase the risks on reliability. Silent data corruptions (SDCs) or silent errors are among the major sources that corrupt HPC execution results. Unlike fail-stop errors, SDCs can be harmful and dangerous in that they cannot be detected by hardware. To remedy this, we propose an online MAchine-learning-based silent data CORruption Detection framework (abbreviated as MACORD) for detecting SDCs in HPC applications. In our study, we comprehensively investigate the prediction ability of a multitude of machine-learning algorithms and enable the detector to automatically select the best-fit algorithms at runtime to adapt to the data dynamics. Because it takes only spatial features (i.e., neighboring data values for each data point in the current time step) into the training data, our learning framework exhibits low memory overhead (less than 1%). Experiments based on real-world scientific applications/benchmarks show that our framework can elevate the detection sensitivity (i.e., recall) up to 99%. Meanwhile the false positive rate is limited to 0.1% in most cases, which is one order of magnitude improvement compared with the latest state-of-the-art spatial technique.

关键词： Training algorithm design and analysis Heuristic algorithms Vegetation Prediction algorithms Machine learning algorithms Mathematical model

来源：评论

学校读者我要写书评

暂无评论

Systematic design of programs with sub-word parallelism

Systematic design of programs with sub-word parallelism

引用

International Conference on Parallel Computing in Electrical Engineering (PAR ELEC 2002)

作者： Schaffer, R Merker, R Catthoor, F Tech Univ Dresden Inst Circuits & Syst Dresden Germany

ISBN: (纸本)0769517307;0769517315

Upcoming processor generations increasingly provide instructions for sub-word parallelism. Thus, a parallel execution of 2, 4 or 8 instructions (add, sub) or of complex instructions (sum of differences) with an input of 2, 4 or 8 operand pairs becomes possible. The exploitation of sub-word parallelism is still weakly supported by current compilers. To close this gap we present an approach to generate progranisfor processors with sub-word parallelism. To this end we adapt methods from the design of parallel processor arrays. An algorithm representing a short term analysis filtering is used to illustrate the approach.

关键词： algorithm design and analysis Circuits and systems design methodology Difference equations Digital signal processing Filtering algorithms Instruction sets Parallel processing Process design Signal processing algorithms

来源：评论

学校读者我要写书评

暂无评论

On strict consistency of a class of stabilizing switching signals for discrete-time switched linear systems

On strict consistency of a class of stabilizing switching si...

引用

European Control Conference (ECC)

作者： Kundu, Atreyee Chatterjee, Debasish Daafouz, Jamal Indian Inst Technol Syst & Control Engn Bombay Maharashtra India Univ Lorraine CRAN CNRS F-54516 Vandoeuvre Les Nancy France Inst Univ France Paris France

ISBN: (纸本)9783952426937

We consider the class of stabilizing switching signals for discrete-time switched linear systems proposed in [9] and study strict consistency property of the same with respect to quadratic cost. Given the underlying digraph of a switched system, we identify circuits on this digraph such that a switching signal corresponding to an infinite walk obtained by repeating the above circuit ensures improved quadratic cost.

关键词： directed graphs discrete time systems linear systems stability switching systems (control) circuits identification digraph discrete-time switched linear systems infinite walk quadratic cost stabilizing switching signals strict consistency property algorithm design and analysis Asymptotic stability Indexes Linear systems Switched systems Switches Switching circuits

来源：评论

学校读者我要写书评

暂无评论

On the use of a high-level fault model to check properties incompleteness 1

On the use of a high-level fault model to check properties i...

引用

1st ACM/IEEE International Conference on Formal Methods and Models for Co-design (MEMOCODE 03)

作者： Fummi, F Pravadelli, G Fedeli, A Rossi, U Toto, F Univ Verona Dipartimento Informat I-37134 Verona Italy

ISBN: (纸本)0769519237

The use of model checking to validate descriptions of digital systems lacks a coverage metrics. The set of proven properties can be incomplete, thus not guaranteeing the behavioral checking completeness of the digital system implementation with respect to the specification. This paper proposes a coverage methodology based on a combination of model checking, high-level fault simulation and automatic test pattern generation, to estimate the incompleteness of a set of formal properties. The adopted high-level fault model allows to join dynamic and formal verification.

关键词： algorithm design and analysis Automatic test pattern generation design engineering Digital systems Explosions Formal verification Logic design Mathematical model Safety State-space methods

来源：评论

学校读者我要写书评

暂无评论

Fast and Exact Monitoring of Co-evolving Data Streams 14

Fast and Exact Monitoring of Co-evolving Data Streams

引用

14th IEEE International Conference on Data Mining (IEEE ICDM)

作者： Matsubara, Yasuko Sakurai, Yasushi Ueda, Naonori Yoshikawa, Masatoshi Kumamoto Univ Kumamoto 860 Japan NTT Commun Sci Labs Kyoto Japan Kyoto Univ Kyoto 6068501 Japan

ISBN: (纸本)9781479943036

Given a huge stream of multiple co-evolving sequences, such as motion capture and web-click logs, how can we find meaningful patterns and spot anomalies? Our aim is to monitor data streams statistically, and find subsequences that have the characteristics of a given hidden Markov model (HMM). For example, consider an online web-click stream, where massive amounts of access logs of millions of users are continuously generated every second. So how can we find meaningful building blocks and typical access patterns such as weekday/weekend patterns, and also, detect anomalies and intrusions? In this paper, we propose StreamScan, a fast and exact algorithm for monitoring multiple co-evolving data streams. Our method has the following advantages: (a) it is effective, leading to novel discoveries and surprising outliers, (b) it is exact, and we theoretically prove that StreamScan guarantees the exactness of the output, (c) it is fast, and requires O(1) time and space per time-tick. Our experiments on 67GB of real data illustrate that StreamScan does indeed detect the qualifying subsequence patterns correctly and that it can offer great improvements in speed (up to 479,000 times) over its competitors.

关键词： computational complexity hidden Markov models media streaming HMM StreamScan access pattern detection co-evolving data stream monitoring hidden Markov model algorithm design and analysis Heuristic algorithms Hidden Markov models Legged locomotion Monitoring Periodic structures Viterbi algorithm Co-evolving data streams

来源：评论

学校读者我要写书评

暂无评论

Cost Optimization based on Brain Storming for Grid Scheduling 4

Cost Optimization based on Brain Storming for Grid Schedulin...

引用

4th International Conference on Innovative Computing Technology (INTECH) / 3rd International Conference on Future Generation Communication Technologies (FGCT)

作者： Arsuaga-Rios, Maria Vega-Rodriguez, Miguel A. CERN European Org Nucl Res IT Dept Geneva Switzerland Univ Extremadura Dept Technol Comp & Commun Escuela Politecn ARCO Res Grp Caceres Spain

ISBN: (纸本)9781479942336

Job scheduling is a challenging task in terms of execution tune, which is usually minimised to accomplish deadlines in diferent areas. However, cost optimization is a conflictive objective when the execution time is optimized, because usually expensive resources are the fastest ones. In this paper, we present a multiobjective algorithm based on the brain storming process that human groups carried out in order to achieve the best initiative or ideas to optimize the job scheduling problem in Grid environments. This swarm algorithm, MOBSA (multiobjective Brain Storm algorithm), minimizes both objectives: execution time and cost without any weights of influence. The standard multiobjective algorithm NSGA-H is compared with MOBSA to evaluate its multiobjective quality. The results obtained by MOBSA are superior than the results obtained by MOHEFT which is the multiobjective version of HEFT, one of the mostused algorithms in workflow scheduling. Furthermore, MOBSA's results also surpass the results provided by real grid schedulers as WMS from gLite middleware or DBC from Nimrod-G.

关键词： algorithm design and analysis Boron Linux Resource management Scheduling Sociology Vectors algorithm design and analysis Sociology Linux execution time management of resources Group problem solving Boron Human groups grids Grid FBN1 wt Allele

来源：评论

学校读者我要写书评

暂无评论

A fast analysis algorithm for 2D multi-reflector quasi-optical system based on Gaussian beam mode analysis 12

A fast analysis algorithm for 2D multi-reflector quasi-optic...

引用

12th UK-Europe-China Workshop on Millimeter Waves and Terahertz Technologies (UCMMT)

作者： Chen, Tianyang Chen, Xiaodong Zheng, Yiyang Yao, Yuan Yu, Junsheng Liu, Xiaoming Beijing Univ Posts & Telecommun Sch Elect Engn Beijing Peoples R China Queen Marry Univ London Sch Elect Engn & Comp Sci London England Anhui Normal Univ Coll Phys & Elect Informat Wuhu Anhui Peoples R China

ISBN: (纸本)9781728129938

In the millimeter and sub-millimeter wave bands, the analysis of electrically large reflectors based on Physical Optics (PO) becomes inefficient. In this paper, the transformation of Gaussian beam modes of co-polarization and cross-polarization in 2D multi-reflector system based on Gaussian beam mode analysis (GBMA) is described in detail and some significant expressions are given and discussed. Then we proposed a fast analysis method based on vector addition method. In order to verify the validity and accuracy of this approach, we compare the results of our algorithm and PO in commercial software GRASP10 with some examples of conic surface mirrors containing the amplitude and phase.

关键词： Millimeter wave and Submillimeter wave technology Quasi-optical system Gaussian beam mode analysis Physical Optics algorithm design and analysis

来源：评论

学校读者我要写书评

暂无评论

Coding of spectral magnitudes using optimized linear transformations

Coding of spectral magnitudes using optimized linear transfo...

引用

7th IEEE Workshop on Speech Coding

作者： Etemoglu, ÇÖ Cuperman, V Univ Calif Santa Barbara Dept Elect & Comp Engn Santa Barbara CA 93106 USA

ISBN: (纸本)0780364163

This paper introduces a novel vector quantization (VQ) technique, wherein the quantized vector is obtained by applying a linear transformation selected from a first codebook to a codevector selected from a second codebook. The transformation is selected from a family of linear transformations, represented by a matrix codebook. Vectors in the second codebook are called residual codevectors. In order to avoid high complexity during the search for the best linear transformation, each linear transformation is assigned a representative vector, such that the search can be done employing the representative vectors, The design algorithm is based on joint optimization of the linear transformation and the residual codebooks. It is shown that the proposed technique yields high quality spectral magnitude quantizer with performance exceeding that of multistage vector quantizer (MSVQ) of similar complexity and bit rate.

关键词： algorithm design and analysis Bit rate Books design optimization Discrete cosine transforms Ear Frequency Sampling methods Speech Vector quantization

来源：评论

学校读者我要写书评

暂无评论

Limits of Trilateration-Based Sensor Placement algorithms 12

Limits of Trilateration-Based Sensor Placement Algorithms

引用

12th IEEE Sensors Applications Symposium (SAS)

作者： Geneve, Lionel Kermorgant, Olivier Laroche, Edouard Univ Strasbourg CNRS ICube Lab Strasbourg France Ecole Cent Nantes LS2N Nantes France

ISBN: (纸本)9781509032020

To localize a mobile vehicle in a predefined area, it is possible to resort to a positioning system using beacons. The questions arising then are: what is the right number of beacons to deploy, and where should they be positioned to accurately localize the target. With a sensor placement algorithm, it is possible to generate a configuration of the beacons that optimize some quality criteria. However, designing the objectives to optimize is not simple, and the solution of the sensor placement algorithm may not always ensures a good localization in practice. This work evaluates the impact of the objectives of a sensor placement algorithm on different localization techniques. We show that criteria based on trilateration are not sufficient anymore when the localization is done with other techniques than trilateration. Indeed, data fusion-based localization algorithms use additional information such as odometry, and are less sensitive to poor coverage or poor beacon configurations than trilateration. Thus, designing new criteria taking into account the dynamics of the vehicle would probably improve further the placements and use less beacons for the same performance.

关键词： Robot sensing systems Optimization Mobile robots algorithm design and analysis Heuristic algorithms Mobile communication

来源：评论

学校读者我要写书评

暂无评论

Nonlinear system characterization and optimization

引用

IRE Transactions on Information Theory 1959年第5期5卷 30-40页

作者： Bose, Amar G. Department of Electrical Engineering and Research Laboratory of Electronics Massachusetts Institute of Technology Cambridge Massachusetts United States

Characterization and optimization of nonlinear systems, is considered from the function space point of view, and implications of this point of view are discussed. Optimization of nonlinear systems is regarded as the problem of mapping the function space of the past of the input onto a line that corresponds to the amplitude of the filter output. It is shown that a system representation that is orthogonal for all possible inputs results from any mapping that partitions this space into nonoverlapping cells. The complication of solving for optimum systems in terms of measured higher-order statistics is circumvented by formulating the problem so that particular statistical measurements directly yield the optimum systems.© 1959 IEEE. All rights reserved.

关键词： Nonlinear systems Particle measurements Space technology Extraterrestrial measurements Gaussian noise Filters Electric variables measurement Loss measurement Noise measurement algorithm design and analysis

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：