检索结果-内蒙古大学图书馆

您好，读者！请登录

内蒙古大学图书馆

首页
概况
党建
资源
服务
科研支持
- 论文收录引用证明
- 科技查新
知识产权
档案馆
帮助

咨询与建议

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

您的常用邮箱：*

您的手机号码：*

问题描述：

当前已输入0个字，您还可以输入200个字

全部搜索
期刊论文
图书
学位论文
标准
纸本馆藏
外文资源发现
数据库导航
超星发现

高级检索

时间限定

出版年份：

文献类型

图书期刊文献学位论文多媒体

馆藏选择

电子馆藏纸本馆藏

核心期刊

全部期刊 SCI 收录期刊 SSCI 收录期刊 EI 收录期刊 CSCD 收录期刊 CSSCI 收录期刊

语言

中文英文

文献类型

期刊文献图书学位论文标准纸本馆藏

帮助

文字说明：

T=题名（书名、题名），A=作者（责任者），K=主题词，P=出版物名称，PU=出版社名称，O=机构（作者单位、学位授予单位、专利申请人），L=中图分类号，C=学科分类号，U=全部字段，Y=年（出版发行年、学位年度、标准发布年）

检索规则说明：

AND代表“并且”；OR代表“或者”；NOT代表“不包含”；(注意必须大写,运算符两边需空一格)

检索范例：

范例一：(K=图书馆学 OR K=情报学) AND A=范并思 AND Y=1982-2016
范例二：P=计算机应用与软件 AND (U=C++ OR U=Basic) NOT K=Visual AND Y=2011-2016

分类表

所选分类

>> <<

限定检索结果

文献类型

4,281 篇 会议
47 篇 期刊文献
3 册 图书

馆藏范围

4,331 篇 电子文献
0 种 纸本馆藏

日期分布

学科分类号

2,432 篇 工学
- 2,168 篇 计算机科学与技术...
- 1,080 篇 软件工程
- 607 篇 电气工程
- 586 篇 信息与通信工程
- 185 篇 仪器科学与技术
- 138 篇 控制科学与工程
- 120 篇 电子科学与技术（可...
- 64 篇 生物工程
- 59 篇 机械工程
- 43 篇 动力工程及工程热...
- 38 篇 交通运输工程
- 31 篇 生物医学工程（可授...
- 27 篇 安全科学与工程
- 26 篇 网络空间安全
- 24 篇 光学工程
- 23 篇 土木工程
- 22 篇 材料科学与工程（可...
- 20 篇 建筑学
- 19 篇 航空宇航科学与技...
418 篇 理学
- 253 篇 数学
- 86 篇 物理学
- 72 篇 生物学
- 51 篇 系统科学
- 42 篇 统计学（可授理学、...
261 篇 管理学
- 179 篇 管理科学与工程(可...
- 97 篇 图书情报与档案管...
- 65 篇 工商管理
35 篇 医学
- 30 篇 临床医学
33 篇 法学
- 19 篇 社会学
18 篇 经济学
- 18 篇 应用经济学
11 篇 农学
7 篇 教育学
7 篇 文学
2 篇 军事学
1 篇 艺术学

主题

543 篇 parallel process...
259 篇 computer archite...
238 篇 computational mo...
207 篇 distributed comp...
186 篇 concurrent compu...
179 篇 computer network...
164 篇 network topology
157 篇 hardware
157 篇 graphics process...
148 篇 routing
130 篇 bandwidth
123 篇 application soft...
120 篇 neural networks
120 篇 protocols
107 篇 scalability
104 篇 multicore proces...
103 篇 throughput
102 篇 algorithm design...
102 篇 kernel
101 篇 computer science

机构

15 篇 institute of inf...
14 篇 institute of com...
14 篇 science and tech...
13 篇 college of compu...
13 篇 school of cyber ...
13 篇 shandong fundame...
12 篇 school of comput...
10 篇 science and tech...
9 篇 univ chinese aca...
9 篇 university of ch...
9 篇 natl univ def te...
9 篇 natl univ def te...
8 篇 shenyang aerospa...
8 篇 beijing univ pos...
8 篇 nanjing univ aer...
8 篇 beijing universi...
7 篇 shanghai jiao to...
7 篇 college of compu...
7 篇 chinese acad sci...
7 篇 national key lab...

作者

17 篇 bagherzadeh nade...
16 篇 kotenko igor
14 篇 j. duato
12 篇 wang yijie
11 篇 daneshtalab maso...
10 篇 ebrahimi masoume...
10 篇 yijie wang
10 篇 nader bagherzade...
10 篇 igor kotenko
10 篇 lexi xu
9 篇 li dongsheng
9 篇 marco aldinucci
9 篇 matsutani hiroki
8 篇 hiroki matsutani
8 篇 masoumeh ebrahim...
8 篇 plosila juha
8 篇 marco danelutto
8 篇 wang huaimin
7 篇 tenhunen hannu
7 篇 xinzhou cheng

语言

4,273 篇 英文
49 篇 其他
11 篇 中文
1 篇 乌克兰文

检索条件"任意字段=Euromicro Conference on Parallel, Distributed and Network-Based Processing"

共 4331 条记录，以下是971-980 订阅

全选清除本页清除全部题录导出标记到"检索档案"

详细简洁

排序：

UNDERSTANDING THE EFFECTS OF DATA parallelISM AND SPARSITY ON NEURAL network TRAINING 9

UNDERSTANDING THE EFFECTS OF DATA PARALLELISM AND SPARSITY O...

引用

9th International conference on Learning Representations, ICLR 2021

作者： Lee, Namhoon Torr, Philip H.S. Ajanthan, Thalaiyasingam Jaggi, Martin University of Oxford United Kingdom Australian National University Australia EPFL Switzerland

We study two factors in neural network training: data parallelism and sparsity;here, data parallelism means processing training data in parallel using distributed systems (or equivalently increasing batch size), so that training can be accelerated;for sparsity, we refer to pruning parameters in a neural network model, so as to reduce computational and memory cost. Despite their promising benefits, however, understanding of their effects on neural network training remains elusive. In this work, we first measure these effects rigorously by conducting extensive experiments while tuning all metaparameters involved in the optimization. As a result, we find across various workloads of data set, network model, and optimization algorithm that there exists a general scaling trend between batch size and number of training steps to convergence for the effect of data parallelism, and further, difficulty of training under sparsity. Then, we develop a theoretical analysis based on the convergence properties of stochastic gradient methods and smoothness of the optimization landscape, which illustrates the observed phenomena precisely and generally, establishing a better account of the effects of data parallelism and sparsity on neural network training. © 2021 ICLR 2021 - 9th International conference on Learning Representations. All rights reserved.

关键词： Stochastic systems

来源：评论

学校读者我要写书评

暂无评论

Delay Optimization of Power Internet of Things based on Edge-Cloud Collaboration

Delay Optimization of Power Internet of Things based on Edge...

引用

International conference on Communication Technology (ICCT)

作者： Julaiti Abulizi Qingsheng Hu Wei Wang State Grid Electronic Power Research Institute Urumqi China School of Information Science and Engineering Southeast University Nanjing China Nanjing Jiruizhidian Technology Co. Ltd. Nanjing China

With the continuous development of the power Internet of Things and the emergence of multi service scenarios, real-time and high-speed process in power data services face more and more challenges. This paper presents an SDN based edge-cloud collaborative method for delay optimization. First, the delay optimization problem is formulated and transformed into a convex optimization problem further. Then a continuously variable coefficient is introduced and its optimal solution is found. Finally a valid task partition strategy that closet to the optimal solution can be found, and the system delay is minimized. Simulation results show that the reasonable loading of edge node tasks to the cloud can effectively reduce the total system delay of power business and alleviate network congestion. Compared with other conventional methods, this edge-cloud collaboration method not only has low system delay, but also makes full use of the control mechanism of SDN to coordinate distributed environment.

关键词： Cloud computing Simulation Collaboration Optimization methods parallel processing Real-time systems Delays

来源：评论

学校读者我要写书评

暂无评论

19th International conference on Algorithms and Architectures for parallel processing, ICA3PP 2019

19th International Conference on Algorithms and Architecture...

引用

19th International conference on Algorithms andArchitectures for parallel processing, ICA3PP 2019

ISBN: (纸本)9783030389604

The proceedings contain 102 papers. The special focus in this conference is on Algorithms and Architectures for parallel processing. The topics include: Utility-based Location Distribution Reverse Auction Incentive Mechanism for Mobile Crowd Sensing network;safeguarding Against Active Routing Attack via Online Learning;reliability Aware Cost Optimization for Memory Constrained Cloud Workflows;null Model and Community Structure in Heterogeneous networks;an Asynchronous Algorithm to Reduce the Number of Data Exchanges;two-Stage Clustering Hot Event Detection Model for Micro-blog on Spark;mobility-Aware Workflow Offloading and Scheduling Strategy for Mobile Edge Computing;HSPP: Load-Balanced and Low-Latency File Partition and Placement Strategy on distributed Heterogeneous Storage with Erasure Coding;adaptive Clustering for Outlier Identification in High-Dimensional Data;Improving the parallelism of CESM on GPU;penguin Search Aware Proactive Application Placement;a Data Uploading Strategy in Vehicular Ad-hoc networks Targeted on Dynamic Topology: Clustering and Cooperation;Cloud Server Load Turning Point Prediction based on Feature Enhanced Multi-task LSTM;Neuron Fault Tolerance Capability based Computation Reuse in DNNs;reliability Enhancement of Neural networks via Neuron-Level Vulnerability Quantization;A Fault Detection Algorithm for Cloud Computing Using QPSO-based Weighted One-Class Support Vector Machine;paraMoC: A parallel Model Checker for Pushdown Systems;FastDRC: Fast and Scalable Genome Compression based on distributed and parallel processing;a parallel Approach to Advantage Actor Critic in Deep Reinforcement Learning;Blockchain-PUF-based Secure Authentication Protocol for Internet of Things;parallel Approach to Sliding Window Sums;selective Velocity distributed Indexing for Continuously Moving Objects Model;a New Bitcoin Address Association Method Using a Two-Level Learner Model.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Efficient Wavefront parallel processing for HEVC CABAC Decoding

Efficient Wavefront Parallel Processing for HEVC CABAC Decod...

引用

euromicro conference on parallel, distributed and network-based processing

作者： P. Habermann B. Juurlink C. C. Chi M. Alvarez-Mesa Embedded Systems Architecture TU Berlin Berlin Germany Spin Digital Video Technologies GmbH Berlin Germany

ISBN: (数字)9781728165820

ISBN: (纸本)9781728165837

Context-based Adaptive Binary Arithmetic Coding (CABAC) is the only compute-intensive task in the High Efficiency Video Coding (HEVC) Standard that does not contain significant data-level parallelism. As a result, it is often a throughput bottleneck for the overall decoding process, especially for high-quality videos. Consequently, the use of high-level parallelization techniques is inevitable to reach throughput requirements for CABAC decoding. Multiple high-level parallelization tools are specified in HEVC, amongst which Wavefront parallel processing (WPP) has only small losses in coding efficiency. However, it lacks in parallel efficiency due to a ramp-up and -down in active parallel threads within a frame. This is a serious problem for systems that cannot process multiple frames at the same time due to performance or memory constraints (e.g. mobile devices), and also for low-delay applications such as video conferencing. To address this issue, we present three improved WPP implementations for HEVC CABAC decoding. They differ in the granularity at which dependency checks are performed. The improvement comes from increased parallel efficiency of the WPP implementation while using the same number of threads as conventional WPP. The proposed implementations allow speedups up to 1.83 × with very little implementation overhead.

关键词： Decoding parallel processing Instruction sets Streaming media Context Context modeling Video coding

来源：评论

学校读者我要写书评

暂无评论

Optimizing FFT-based Convolution on ARMv8 Multi-core CPUs 26th

Optimizing FFT-Based Convolution on ARMv8 Multi-core CPUs

引用

26th International conference on parallel and distributed Computing (Euro-Par)

作者： Wang, Qinglin Li, Dongsheng Huang, Xiandong Shen, Siqi Mei, Songzhu Liu, Jie Natl Univ Def Technol Sci & Technol Parallel & Distributed Proc Lab Changsha 410073 Peoples R China Natl Univ Def Technol Coll Comp Changsha 410073 Peoples R China

ISBN: (纸本)9783030576752;9783030576745

Convolutional Neural networks (CNNs) are widely applied in various machine learning applications and very time-consuming. Most of CNNs' execution time is consumed by convolutional layers. A common approach to implementing convolutions is the FFT-based one, which can reduce the arithmetic complexity of convolutions without losing too much precision. As the performance of ARMv8 multi-core CPUs improves, they can also be utilized to perform CNNs like Intel X86 CPUs. In this paper, we present a new parallel FFT-based convolution implementation on ARMv8 multi-core CPUs. The implementation makes efficient use of ARMv8 multi-core CPUs through a series of computation and memory optimizations. The experiment results on two ARMv8 multicore CPUs demonstrate that our new implementation gives much better performance than two existing approaches in most cases.

关键词： CNNs Convolution FFT ARMv8 parallel algorithm

来源：评论

学校读者我要写书评

暂无评论

NetANNS: A High-Performance distributed Search Framework based On In-network Computing

NetANNS: A High-Performance Distributed Search Framework Bas...

引用

IEEE International conference on Big Data and Cloud Computing (BdCloud)

作者： Penghao Zhang Heng Pan Zhenyu Li Gaogang Xie Penglai Cui Institute of Computing Technology Chinese Academy of Sciences China University of Chinese Academy of Sciences China Purple Mountain Laboratories China Computer Network Information Center Chinese Academy of Sciences China

Approximate nearest neighbor search (ANNS) is the most basic and important algorithm in Database, Machine Learning and other applications. With the expansion of cloud computing, the academia focuses on the study of how to optimize distributed frameworks based on approximate nearest neighbor search such as MapReduce, and Memcached. We implement a new distributed ANNS search framework (NetANNS). The main contributions of NetANNS are to accelerate the data preprocessing with programmable switch, and integrate a variety of efficient ANNS algorithms so that it can choose the most suitable algorithm for each datasets. The experiments show that the search efficiency of NetANNS is about 2x than the common distributed ANNS frameworks which are implemented based on the framework of MapReduce.

关键词： Cloud computing Machine learning algorithms Data preprocessing Switches Nearest neighbor methods Approximation algorithms Database machines

来源：评论

学校读者我要写书评

暂无评论

Selective Caching: Avoiding Performance Valleys in Massively parallel Architectures

Selective Caching: Avoiding Performance Valleys in Massively...

引用

euromicro conference on parallel, distributed and network-based processing

作者： Amin Jadidi Mahmut T. Kandemir Chita R. Das Cadence Design Systems Department of Computer Science and Engineering Pennsylvania State University USA

ISBN: (数字)9781728165820

ISBN: (纸本)9781728165837

Emerging general purpose graphics processing units (GPGPU) make use of a memory hierarchy very similar to that of modern multi-core processors - they typically have multiple levels of on-chip caches and a DDR-like off-chip main memory. In such massively parallel architectures, caches are expected to reduce the average data access latency by reducing the number of off-chip memory accesses; however, our extensive experimental studies confirm that not all applications utilize the on-chip caches in an efficient manner. Even though GPGPUs are adopted to run a wide range of general purpose applications, the conventional cache management policies are incapable of achieving the optimal performance over different memory characteristics of the applications. This paper first investigates the underlying reasons for inefficiency of common cache management policies in GPGPUs. To address and resolve those issues, we then propose (i) a characterization mechanism to analyze each kernel at runtime and, (ii) a selective caching policy to manage the flow of cache accesses. Evaluation results of the studied platform show that our proposed dynamically reconfigurable cache hierarchy improves the system performance by up to 105% (average of 27%) over a wide range of modern GPGPU applications, which is within 10% of the optimal improvement.

关键词： Kernel Instruction sets Graphics processing units System performance System-on-chip parallel processing Delays

来源：评论

学校读者我要写书评

暂无评论

A Memory Saving Mechanism based on Data Transferring for Pipeline parallelism

A Memory Saving Mechanism Based on Data Transferring for Pip...

引用

IEEE International conference on Big Data and Cloud Computing (BdCloud)

作者： Wei Jiang Rui Xu Sheng Ma Qiong Wang Xiang Hou Hongyi Lu Science and Technology on Parallel and Distributed Processing Laboratory College of Computer National University of Defense Technology Changsha China

The correctness and robustness of the neural network model are usually proportional to its depth and width. Currently, the neural network models become deeper and wider to cope with complex applications, which leads to high memory capacity requirement and computer capacity requirements of the training process. The multi-accelerator parallelism is a promising choice for the two challenges, which deploys multiple accelerators in parallel for training neural networks. Among them, the pipeline parallel scheme has a great advantage in training speed, but its memory capacity requirements are relatively higher than other parallel schemes. Aiming at solving this challenge of pipeline parallel scheme, we propose a data transfer mechanism, which effectively reduces the peak memory usage of the training process by real-time data transferring. In the experiment, we implement our design and apply it to Pipedream, a mature pipeline parallel scheme. The memory requirement of training process is reduced by up to 48.5%, and the speed loss is kept within a reasonable range.

关键词： Training Computational modeling Pipelines Neural networks Memory management Bandwidth parallel processing

来源：评论

学校读者我要写书评

暂无评论

distributed Root-MUSIC Using Finite-Time Average Consensus

Distributed Root-MUSIC Using Finite-Time Average Consensus

引用

Asilomar conference on Signals, Systems & Computers

作者： Po-Chih Chen P. P. Vaidyanathan 136-93 California Institute of Technology Pasadena CA USA

distributed (decentralized) algorithms for principal component analysis of covariance matrices are well-known, and their applications in array signal processing have received more interest recently. Inspired by this, a new distributed algorithm for the popular DOA estimation method, root-MUSIC, is proposed. The average consensus method is used to avoid the need for a fusion center in a sensor network, which makes the proposed algorithm fully distributed. In particular, the algorithm is based on a recently reported finite-time version of average consensus which converges to the exact solution in a finite number of iterations. This allows the proposed distributed root-MUSIC to achieve exactly the same performance as the centralized counterpart. The good performance of the proposed algorithm is verified by simulations.

关键词： Computers Direction-of-arrival estimation Array signal processing Signal processing algorithms Estimation parallel processing distributed algorithms

来源：评论

学校读者我要写书评

暂无评论

Gene Regulatory network Construction parallel Technique based on network Component Analysis 1st

Gene Regulatory Network Construction Parallel Technique Base...

引用

1st International conference on Artificial Intelligence and Computer Visions, AICV 2020

作者： Elsayad, Dina Hamad, Safwat Shedeed, Howida A. Tolba, Mohamed F. Faculty of Computer and Information Sciences Ain Shams University Cairo Egypt

ISBN: (纸本)9783030442880

Construction of a gene regulatory network is a vital process for understanding the gene functions and gene influences on the other genes. Furthermore, gene regulatory network analysis is a promising method for demonstrating the topological order of genes interactions. One technique for constructing a gene regulatory network is FastNCA that is based on network component analysis methodology. Although FastNCA is widely used to construct the gene regulatory network of cancer diseases associated genes, it is a time-consuming and computational intensive technique. As a result, this paper presents an enhanced parallel implementation of FastNCA that uses distributed parallelism methodology to enhance the performance of FastNCA. Different gene datasets are used to evaluate the performance of the proposed algorithm. The experimental results demonstrate that the proposed algorithm outperforms FastNCA. Where the achieved speedup is up to 250 on 256 processing nodes. © 2020, Springer Nature Switzerland AG.

关键词： Genes

来源：评论

学校读者我要写书评

暂无评论

没有更多数据了...

全选清除本页清除全部题录导出标记到“检索档案”

共434页 << < 94 95 96 97 98 99 100 101 102 103 > >>

检索报告对象比较合并检索0

隐藏清空

合并搜索

回到顶部

执行限定条件

内容：

评分：

请选择保存的检索档案：

请选择收藏分类：

订阅名称：

通借通还

温馨提示：

图书名称：

借书校区：

取书校区：

手机号码：

邮箱地址：

一卡通帐号：

电话和邮箱必须正确填写，我们会与您联系确认。

联系人：

所在院系：

联系邮箱：

联系电话：

内蒙古自治区呼和浩特市赛罕区大学西街235号邮编: 010021

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：