检索结果-内蒙古大学图书馆

arXiv 2018年

作者： Shen, Junzhong Qiao, Yuran Huang, You Wen, Mei Zhang, Chunyuan College of Computer National University of Defense Technology Changsha410073 China National Key Laboratory for Parallel and Distributed Processing National University of Defense Technology Changsha410073 China

Large-scale floating-point matrix multiplication is a fundamental kernel in many scientific and engineering applications. Most existing work only focus on accelerating matrix multiplication on FPGA by adopting a linear systolic array. This paper towards the extension of this architecture by proposing a scalable and highly configurable multi-array architecture. In addition, we propose a work-stealing scheme to ensure the equality in the workload partition among multiple linear arrays. Furthermore, an analytical model is developed to determine the optimal design parameters. Experiments on a real-life convolutional neural network (CNN) show that we can obtain the optimal extension of the linear array architecture. Copyright © 2018, The Authors. All rights reserved.

关键词： Field programmable gate arrays (FPGA)

来源：评论

学校读者我要写书评

暂无评论

Implementation of encrypted data for outsourced database

Implementation of encrypted data for outsourced database

引用

International Conference on Computational Intelligence and Natural Computing, CINC

作者： Zheng-Fei Wang Ai-Guo Tang National Key Laboratory of Parallel and Distributed Processing School of Computer National University of Defense Technology Changsha China Department of Computer Hunan Business College Changsha China

ISBN: (纸本)9781424477050

Encryption technology has become an important mechanism of securing data stored in the outsourced database. However, it is a difficulty to query efficiently the encrypted data and many researchers take it into consideration. To solve the problem, an encrypted schema, based on the Postgresql DBMS, is proposed. Through the security dictionary and the extended SQL, the approach implements the encrypted storage and efficiently query over the encrypted data in the outsourced databases. Results of experiments validate the efficiency and feasibility of our approach.

关键词： Indexes Dictionaries Random access memory Catalogs Encryption

来源：评论

学校读者我要写书评

暂无评论

Detecting faults by tracing companion states in cloud computing systems

引用

Jisuanji Xuebao/Chinese Journal of Computers 2012年第5期35卷 856-870页

作者： Rao, Xiang Wang, Huai-Min Chen, Zhen-Bang Zhou, Yang-Fan Cai, Hua Zhou, Qi Sun, Ting-Tao National Key Laboratory for Parallel and Distributed Processing National University of Defense Technology Changsha 410073 China Shenzhen Research Institute The Chinese University of Hong Kong Shenzhen China Department of Computing Platform Alibaba Cloud Computing Company Hangzhou 310011 China

A common way to construct a fault model is injecting the fault into the system and observing the subsequent symptoms, e. g. event logs. However, fault features would vary during the propagation period, and present different symptoms at different stage of the fault propagation process. The exiting detection window based feature extraction methods can only identify the early symptoms of a fault, but fail to detect the latter symptoms and cause false alarms. To solve the problem, we present a fault feature extraction method, called Companion State Tracer (CSTracer), which consists of 3 integrated steps: (1) pre-process logs to remove the unrelated logs;(2) construct a general identifier for the early symptoms of a fault;(3) construct a finite state machine model for the fault to trace the latter symptoms. CSTracer can persistently monitor a fault after the fault has been identified. We have justified the effectiveness of CSTracer in an enterprise cloud system. Compared with the existing, the results show that CSTracer has a better detection accuracy.

关键词： Fault detection

来源：评论

学校读者我要写书评

暂无评论

Predicting Programming Behavior in OSS Communities: A Case Study of NLP-based Approach

Predicting Programming Behavior in OSS Communities: A Case S...

引用

Artificial Intelligence and Computer Engineering (ICAICE), International Conference on

作者： Manyan Huo Yue Yu Zhixing Li Junsheng Chang Key Lab. of Parallel and Distributed Computing Lab. College of Computer National University of Defence Technology ChangSha China

ISBN: (数字)9781728191461

ISBN: (纸本)9781728191478

Prediction of developers' programming behaviors is an effective way to improve their development efficiency and optimize the organization of project modules and files. However, little research exists investigating on this direction. In order to address this knowledge gap, we proposed a NLP-based approach to predict the programming behaviors in OSS (Open Source Software) communities. The proposed approach i) embeds the historical programming behavior data of a project into a multi-dimensional vector space to capture the potential laws in the data, ii) forms an eigenvector matrix reflecting the semantic relationship of the development behavior data, and iii) predicts the next programming behavior of a specific developer based on the eigenvector matrix. Our experiments on five OSS projects show that the prediction accuracy rate of the proposed prediction approach can reach up to about 50%, indicating that it can summarize the development behavior data law and effectively predict the programming behavior of developers. Our work can provide valuable assistance for developers' programming and projects' maintenance in practice.

关键词： Training Semantics Programming Predictive models Prediction algorithms Data models Context modeling

来源：评论

学校读者我要写书评

暂无评论

EmSBoT: A lightweight modular software framework for networked robotic systems

EmSBoT: A lightweight modular software framework for network...

引用

International Conference on Advances in Computational Tools for Engineering Applications, ACTEA

作者： Long Peng Fei Guan Luc Perneel Hasan Fayyad-Kazan Martin Timmerman Department of Electronics and Informatics Vrije Universiteit Brussel (VUB) Brussel-Belgium National Key Laboratory of Parallel and Distributed Processing National University of Defense Technology Changsha China

ISBN: (纸本)9781467385244

Developing applications for modern complex networked robotic systems is more challenging due to the introduction of possibly sophisticated communication and coordination aspects. In this paper, we propose EmSBoT, a lightweight embedded component-based software framework targeting resource-constrained networked robotic systems. EmSBoT provides a unified Application Program Interface (API) that hides the heterogeneous distributed environment from applications. Its OS abstraction layer endows it with OS independence and portability. A port-based communication mechanism is adopted to exchange message between loosely coupled components, making the system with fault-tolerance capability. By isolating the communication channels as separate agents, the framework provides uniform and transparent message-passing for agents over node boundaries. We describe the architecture, programming model and core features of EmSBoT in this paper, together with the performance evaluation and behavior validation to demonstrate its efficiency and feasibility.

关键词： Ports (Computers) Robot kinematics Real-time systems Programming Operating systems

来源：评论

学校读者我要写书评

暂无评论

Towards MLOps: A Case Study of ML Pipeline Platform

Towards MLOps: A Case Study of ML Pipeline Platform

引用

Artificial Intelligence and Computer Engineering (ICAICE), International Conference on

作者： Yue Zhou Yue Yu Bo Ding Key Lab. of Parallel and Distributed Computing Lab. College of Computer National University of Defence Technology ChangSha China

ISBN: (数字)9781728191461

ISBN: (纸本)9781728191478

The development and deployment of machine learning (ML) applications differ significantly from traditional applications in many ways, which have led to an increasing need for efficient and reliable production of ML applications and supported infrastructures. Though platforms such as TensorFlow Extended (TFX), ModelOps, and Kubeflow have provided end-to-end lifecycle management for ML applications by orchestrating its phases into multistep ML pipelines, their performance is still uncertain. To address this, we built a functional ML platform with DevOps capability from existing continuous integration (CI) or continuous delivery (CD) tools and Kubeflow, constructed and ran ML pipelines to train models with different layers and hyperparameters while time and computing resources consumed were recorded. On this basis, we analyzed the time and resource consumption of each step in the ML pipeline, explored the consumption concerning the ML platform and computational models, and proposed potential performance bottlenecks such as GPU utilization. Our work provides a valuable reference for ML pipeline platform construction in practice.

关键词： Training Computational modeling Pipelines Graphics processing units Tools Data models Task analysis

来源：评论

学校读者我要写书评

暂无评论

Server-uncertain spectrum trading in cognitive radio networks: A queueing-theoretic modeling approach

Server-uncertain spectrum trading in cognitive radio network...

引用

International Conference on Cognitive Radio Oriented Wireless Networks and Communications

作者： Lixia Liu Gang Hu Ming Xu Yuxing Peng National Key Laboratory for Parallel and Distributed Processing National University of Defense Technology Changsha China Department of Network Engineering College of Computer National University of Defense Technology Changsha China

Spectrum trading is the promising method to improve spectrum efficiency from the perspective of economics. In this paper we propose a queueing-theory based spectrum trading model, where the primary user plays the server role providing spectrum to the secondary user who acts as the customer. The most significant challenge is how to optimize the spectrum trading model considering the server uncertainty which includes service state, service time, service area, service content and service price. We design a STACP queueing model according to the server attributes, so that the secondary user can choose the right queue quickly and reasonably according to its demand. Moreover, we further analyze the optimizing strategies for STACP model which can maximize the profit of the primary user and minimize the service cost of the secondary user. The simulation results demonstrate the analysis results.

关键词： Servers Uncertainty Loss measurement Queueing analysis Economics Analytical models Business

来源：评论

学校读者我要写书评

暂无评论

Grid computing in China

引用

Journal of Grid Computing 2004年第2期2卷 193-206页

作者： Yang, Guangwen Jin, Hai Li, Minglu Xiao, Nong Li, Wei Wu, Zhaohui Wu, Yongwei Tang, Feilong Department of Computer Science and Technology Tsinghua University Beijing China School of Computer Huazhong University of Science and Technology Wuhan China Department of Computer Science and Engineering Shanghai Jiao Tong University Shanghai China Key Laboratory for Parallel and Distributed Processing National University of Defense Technology Changsha China Institute of Computing Technology Chinese Academy of Sciences Beijing 100080 China College of Computer Science Zhejiang University Hangzhou 310027 China

Grid computing presents a new trend to distributed computation and Internet applications, which can construct a virtual single image of heterogeneous resources, provide uniform application interface and integrate widespread computational resources into super, ubiquitous and transparent aggregation. In the adoption of Grid computing, China, who is facing more resource heterogeneity and other specific demands, has put much effort on both research and practical utilization. In this paper, we introduce the major China Grid research projects and their perspective applications. First we give the overview of the four government-sponsored programs in Grid, namely the China national Grid, ChinaGrid, NSFC Grid, and ShanghaiGrid. Then we present six representative ongoing Grid systems in details, which are categorized into Grid middleware and Grid application. This paper provides the general picture of Grid computing in China, and shows the great efforts, devotion and confidence in China to use Grid technology to boost the society, economics and scientific research. © Springer 2004.

关键词： Grid computing

来源：评论

学校读者我要写书评

暂无评论

Attention-based fault-tolerant approach for multi-agent reinforcement learning systems

arXiv

引用

arXiv 2019年

作者： Geng, Mingyang Xu, Kele Li, Yiying Liu, Shuqi Ding, Bo Wang, Huaimin National Key Laboratory of Parallel and Distributed Processing College of Computer National University of Defense Technology Changsha410073 China National Key Laboratory of Big Data Management and Analysis Northeastern University Shenyang110000 China

The aim of multi-agent reinforcement learning systems is to provide interacting agents with the ability to collaboratively learn and adapt to the behavior of other agents. In many real-world applications, the agents can only acquire a partial view of the world. However, in realistic settings, one or more agents that show arbitrarily faulty or malicious behavior may suffice to let the current coordination mechanisms fail. In this paper, we study a practical scenario considering the security issues in the presence of agents with arbitrarily faulty or malicious behavior. Under these circumstances, learning an optimal policy becomes particularly challenging, even in the unrealistic case that an agent's policy can be made conditional upon all other agents' observations. To overcome these difficulties, we present an Attention-based Fault-Tolerant (FT-Attn) algorithm which selects correct and relevant information for each agent at every time-step. The multi-head attention mechanism enables the agents to learn effective communication policies through experience concurrently to the action policies. Empirical results have shown that FT-Attn beats previous state-of-the-art methods in some complex environments and can adapt to various kinds of noisy environments without tuning the complexity of the algorithm. Furthermore, FT-Attn can effectively deal with the complex situation where an agent needs to reach multiple agents' correct observation at the same time. Copyright © 2019, The Authors. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

A data-driven mechanism for large-scale data distribution

A data-driven mechanism for large-scale data distribution

引用

Proceedings of the Biannual World Automation Congress

作者： Peichang Shi Yiying Li Bo Ding Longquan Jiang Hui Liu Jie Zhang National Key Laboratory of Parallel and Distributed Processing School of Computer Science National University of Defense Technology Changsha Hunan CN National Key Laboratory of Parallel and Distributed Processing School of Computer Science National University of Defense Technology Changsha 410073 China National University of Defense Technology Changsha Hunan CN China Electr. Equip. & Syst. Eng. Co. Ltd. China

As The integration of Physical space and cyberspace, the large-scale data distributing to diversification terminal which is geographical distribution of mass has become a huge challenge. When the data size can't be processed by the technology for traditional scope, how to deal with the user quality of service and efficient use of system resources has become an important issue of concern, with the resources becoming limited. This paper presents a data-driven mechanism for large-scale data distribution which is consists of four core part of the data production, data collection and pre-processing, data analysis engine, data consumption, aims to excavate the valuable information to improve the efficiency of resource use and accurate fault location for the Large-scale data distribution system. At the same time, this paper studies the resource scheduling optimization with analyzing data driven for the system behavior and Fault location with analyzing data-driven environment, which proves the effectiveness for the operation of the Large-scale data distribution system optimization by the data-driven working.

关键词： Servers Monitoring distributed databases Big data Real-time systems Business

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：