检索结果-内蒙古大学图书馆

International conference on parallel and distributed Systems (ICPADS)

作者： Shanshan Duan Weidong Yang Xuyu Wang Shiwen Mao Yuan Zhang Henan Key Laboratory of Grain Photoelectric Detection & Control Key Laboratory of Grain Information Processing & Control College of Information Science & Engineering Henan University of Technology Zhengzhou Henan China California State University Sacramento CA USA Auburn University Auburn AL USA

ISBN: (数字)9781728190747

ISBN: (纸本)9781728183824

Temperature is one of the major ecological factors that affect the safe storage of grain. In this paper, we propose a deep spatio-temporal attention mode to predict stored grain temperature, which exploits the historical temperature data of stored grain and the meteorological data of the region. In this proposed model, we use the Sobel operator to extract the local spatial factors, and leverage the attention mechanism to obtain the global spatial factors of grain temperature data and temporal information. In addition, a convolutional neural network (CNN) is used to learn features of external meteorological factors. Finally, the spatial factors of grain pile and external meteorological factors are combined to predict future grain temperature using long short-term memory (LSTM) based encoder and decoder models. Experiment results show that the proposed model achieves higher predication accuracy compared with the traditional methods.

关键词： Temperature distribution Meteorological factors Biological system modeling Predictive models Feature extraction Data models Forecasting

来源：评论

学校读者我要写书评

暂无评论

A multi-threaded approach to encoding with random linear network coding 25

A multi-threaded approach to encoding with random linear net...

引用

25th European Wireless conference 2019: Wireless Futures in the Era of network Programmability, EW 2019

作者： Nielsen, Lars Hansen, René Rydhof Lucani, Daniel E. Department of Engineering DIGIT Aarhus University Aarhus Denmark Department of Computer Science Aalborg University Aalborg Denmark

ISBN: (纸本)9783800749492

Cloud and distributed storage applications require processing of large fragments of data. This poses memory, delay, and processing speed challenges for systems using erasure codes to reduce the cost of storage and/or increase the reliability of the system. To address these, this paper proposes and deploys designs that exploit current multi-threading capabilities of microprocessors to accelerate the encoding and decoding process of erasure codes, focusing on the case of Random Linear network Coding (RLNC). More specifically, we propose a strategy for parallel computation based on splitting symbols into significantly smaller fragments and reordering data to speed up computation and decreasing the potential for thread blocking. We implement the strategy in C++ and carry out benchmark experiments and compare them with the single threaded state-of-the-art block RLNC encoders. We show a reduction of processing time by a factor of three by using our strategies for files of 32 MB or more as well as reducing memory usage drastically in the system. We also show that our approach provides a better scaling than the single-threaded option when increasing the number of fragments that a file is broken into. In other words, distributed storage systems can split files or data into a larger number of fragments without experiencing a speed penalty. © VDE VERLAG GMBH.

关键词： network coding

来源：评论

学校读者我要写书评

暂无评论

A Many-Core Accelerator Design for On-Chip Deep Reinforcement Learning

A Many-Core Accelerator Design for On-Chip Deep Reinforcemen...

引用

IEEE International conference on Computer-Aided Design

作者： Ying Wang Mengdi Wang Bing Li Huawei Li Xiaowei Li State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences Beijing P.R. China Capital Normal University Beijing P.R. China

ISBN: (数字)9781665423243

Deep Reinforcement Learning (DRL) is substantially resource-consuming, and it requires large-scale distributed computing-nodes to learn complicated tasks, like video-game and Go play. This work attempts to down-scale a distributed DRL system into a specialized many-core chip and achieve energy-efficient on-chip DRL. With the customized network-on-Chip that handles the communication of on-chip data and control-signals, we proposed a Synchronous Asynchronous RL Architecture (SARLA) and the according many-core chip that completely avoids the unnecessary data duplication and synchronization activities in multi-node RL systems. In evaluation, the SARLA system achieves considerable energy-efficiency boost over the GPU-based implementations for typical DRL workloads built with OpenAI-gym.

关键词： System-on-chip Computer architecture Reinforcement learning Neural networks Servers parallel processing Training

来源：评论

学校读者我要写书评

暂无评论

GENIE: QoS-guided Dynamic Scheduling for CNN-based Tasks on SME Clusters 22

GENIE: QoS-guided Dynamic Scheduling for CNN-based Tasks on ...

引用

22nd Design, Automation and Test in Europe conference and Exhibition (DATE)

作者： Chen, Zhaoyun Luo, Lei Yang, Haoduo Yu, Jie Wen, Mei Zhang, Chunyuan Natl Univ Def Technol Coll Comp Changsha Hunan Peoples R China Natl Key Lab Parallel & Distributed Proc Changsha Hunan Peoples R China

ISBN: (纸本)9783981926323

Convolutional Neural network (CNN) has achieved dramatic developments in emerging Machine Learning (ML) services. Compared to online ML services, offline ML services that are full of diverse CNN workloads are common in small and medium-sized enterprises (SMEs), research institutes and universities. Efficient scheduling and processing of multiple CNN-based tasks on SME clusters is both significant and challenging. Existing schedulers cannot predict the resource requirements of CNN-based tasks. In this paper, we propose GENIE, a QoS-guided dynamic scheduling framework for SME clusters that achieves users' QoS guarantee and high system utilization. based on a prediction model derived from lightweight profiling, a QoS-guided scheduling strategy is proposed to identify the best placements for CNN-based tasks. We implement GENIE as a plugin of Tensorflow and experiment with real SME clusters and large-scale simulations. The results of the experiments demonstrate that the QoS-guided strategy outperforms other baseline schedulers by up to 67.4% and 28.2% in terms of QoS-guarantee percentage and makespan.

关键词： QoS-guided SME Cluster Scheduling CNN Multi-task

来源：评论

学校读者我要写书评

暂无评论

SPAR-2: A SIMD Processor Array for Machine Learning in IoT Devices

SPAR-2: A SIMD Processor Array for Machine Learning in IoT D...

引用

International conference on Data Intelligence and Security (ICDIS)

作者： Suhail Basalama Atiyehsadat Panahi Ange-Thierry Ishimwe David Andrews University of Arkansas Fayetteville Arkansas

ISBN: (数字)9781728193793

ISBN: (纸本)9781728193809

In this paper, SPAR-2, a SIMD processor array for migrating machine learning applications into FPGA IoT edge devices is presented. SPAR-2 is a second-generation processorin/ near-memory architecture developed as a programmable overlay for modern FPGAs. In contrast to point designs, SPAR-2 can be programmed to implement different classes of machine learning algorithms. The overlay architecture is based on Processor-In-Memory (PIM) tiles, which integrate bit-serial ALUs with distributed Block RAMs (BRAMs). Forming PIM tiles increases the size of the multiply-accumulate array and on-chip storage capacity. User-visible inference latencies are reduced by exploiting concurrent accesses to weights and partial results stored throughout the distributed BRAMs. The sizing and performance analysis of SPAR-2 running a standard Long Short-Term Memory (LSTM) Recurrent Neural network (RNN) model is provided. Results show that our approach enables packing up to 16,384 processing elements in a Virtex-7 FPGA. Runtime performance comparisons show that SPAR-2 achieves up to 24.51x speedup compared to a High-Level Synthesis (HLS) equivalent design and 1.75x speedup over a previous custom-tuned design.

关键词： Field programmable gate arrays Registers parallel processing Computer architecture Random access memory Standards Arrays

来源：评论

学校读者我要写书评

暂无评论

network Public Opinion Monitoring based on MapReduce

引用

Journal of Physics: conference Series 2021年第1期2037卷

作者： Yunhong Zhao Suruo Li Jingchu University of Technology Jingmen HuBei China

Public opinion information on the network is abundantly decentralized in Internet, these false ones, easily misleading, not legally binding, it is difficult to deal with. Quickly master information is a prerequisite to deal with the network of public opinion, clustering algorithm in data mining plays an important role in in the collection of information statistics. Disadvantages of classical clustering methods are high resources consumption, low efficiency, high time and space complexity, and difficult to deal with large-scale data processing in massive network. To the network public opinion text as the research object, in-depth study of summarizing the network public opinion monitoring technology based on distributed MapReduce, including parallel technology, improved clustering algorithm, relational database and distributed database construction, which is in order to improve the efficiency of information processing, to reduce the pin.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A parallel Multi-keyword Top-k Search Scheme over Encrypted Cloud Data 16th

A Parallel Multi-keyword Top-k Search Scheme over Encrypted ...

引用

16th IFIP WG 10.3 International conference on network and parallel Computing, NPC 2019

作者： Yang, Maohu Dai, Hua Bao, Jingjing Yi, Xun Yang, Geng Nanjing University of Posts and Telecommunications Nanjing210023 China Jiangsu Security and Intelligent Processing Lab of Big Data Nanjing210023 China Royal Melbourne Institute of Technology University Melbourne3001 Australia

ISBN: (纸本)9783030307080

With searchable encryptions in the cloud computing, users can outsource their sensitive data in ciphertext to the cloud that provides efficient and privacy-preserving multi-keyword top-k searches. However, most existing top-k search schemes over encrypted cloud data are the centralize schemes which are limited in large scale data environment. To support scalable searches, we propose a parallel multi-keyword top-k search scheme over encrypted cloud data. In this scheme, the fragment-based encrypted inverted index is designed, which is indistinguishable and can be used for parallel searching. On the basis of such indexes, a Map-Reduce-based distributed computing framework is adopted to implement the parallel multi-keyword top-k search algorithms. Security analysis and experiment evaluation show that the proposed scheme is privacy-preserving, efficient and scalable. © 2019, IFIP International Federation for Information processing.

关键词： Cloud computing

来源：评论

学校读者我要写书评

暂无评论

Attention-based supply-demand prediction for autonomous vehicles 20

Attention-based supply-demand prediction for autonomous vehi...

引用

20th International conference on parallel and distributed Computing, Applications and Technologies, PDCAT 2019

作者： Zhang, Zikai Dong, Hairong Li, Yidong You, Yizhe Zhao, Fengping School of Electronic and Information Engineering Beijing Jiaotong University Beijing100044 China School of Computer and Information Technology Beijing Jiaotong University Beijing100044 China

ISBN: (纸本)9781728126166

As one of the important functions of the intelligent transportation system (ITS), supply-demand prediction for autonomous vehicles provides a decision basis for its control. In this paper, we present two prediction models (i.e. ARLP model and Advanced ARLP model) based on two system environments that only the current day's historical data is available or several days' historical data are available. These two models jointly consider the spatial, temporal, and semantic relations. Spatial dependency is captured with residual network and dimension reduction. Short term temporal dependency is captured with LSTM. Long term temporal dependency and temporal shifting are captured with LSTM and attention mechanism. Semantic dependency is captured with multi-attention mechanism. Extensive experiments show that our frameworks provide more accurate prediction results than the existing methods. © 2019 IEEE.

关键词： Autonomous vehicles

来源：评论

学校读者我要写书评

暂无评论

Towards workload-adaptive scheduling for HPC clusters

Towards workload-adaptive scheduling for HPC clusters

引用

IEEE International conference on Cluster Computing

作者： Alexander V. Goponenko Ramin Izadpanah Jim M. Brandt Damian Dechev 211 Harris Center (Building 116) Orlando FL Sandia National Laboratories Albuquerque MS

ISBN: (数字)9781728166773

ISBN: (纸本)9781728166780

The performance of HPC clusters depends on efficient scheduling of jobs. However, modern schedulers generally lack real-time information about resource utilization and require users to provide information, which is seldom accurate, on job requirements. The problem is exacerbated as HPC systems become increasingly more complicated and heterogeneous, which gives rise to new resource constraints (GPU, parallel file system, network bandwidth, burst buffers, etc.) In this work, we integrated data from LDMS, the Lightweight distributed Metric Service, with Slurm, a popular job scheduler. To demonstrate the capabilities of such integration, we enabled scheduling based on the Lustre file system throughput. We demonstrated benefits of measurement of real-time utilization, prediction of applications requirements from historical data, and finer control of resources, in a preliminary evaluation of scheduling on a cluster of virtual machines. We also identified the possibility of further improving the scheduling efficiency through workload-adaptive scheduling, by adjusting the scheduling based on characteristics of the pending job. We validated the feasibility of this strategy by simulating job executions in our custom-made HPC cluster simulator.

关键词： Measurement File systems Graphics processing units distributed databases Throughput Real-time systems Resource management

来源：评论

学校读者我要写书评

暂无评论

A Graph Theory-based Clustering Method for Power System networks

A Graph Theory-Based Clustering Method for Power System Netw...

引用

Clemson University Power Systems conference (PSC)

作者： Iroshani Jayawardene Pramod Herath Ganesh Kumar Venayagamoorthy Real-Time Power and Intelligent Systems Laboratory Clemson University Clemson SC USA School of Engineering University of KwaZulu-Natal Durban South Africa

ISBN: (数字)9781728193847

ISBN: (纸本)9781728193854

With the increasing human population and requirements, the current electric power grid is becoming complex and expanding rapidly. Increasing integration of renewable energy sources and other advanced metering technologies introduces a complex, expanded and distributed power system infrastructure. Real-time operation and control of large-scale power system networks have become a challenging task due to these reasons. Power system clustering is a feasible solution to overcome these challenges. Clustering utilizes parallel processing capabilities of control methods and provides adequate methods for real-time power system control. In this paper, a graph theory-based clustering approach is presented for large interconnected power system networks. The algorithm is tested on multiple power system networks from small to large scale and the performance is analyzed. Typical results indicate that the method can provide clusters with uniform sizes within a short time limit for any type of power system network. The clusters observed are applied in solving optimal power flow (OPF) problem.

关键词：

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：