检索结果-内蒙古大学图书馆

Springer International Conference on Emerging Technologies in Data Mining and Information Security (IEMIS)

作者： Prabhu, B. V. Balaji Dakshayini, M. BMS Coll Engn Dept ISE Bangalore 560019 Karnataka India

ISBN: (纸本)9789811319518;9789811319501

Prediction plays an important role everywhere particularly in business, technology, and many others. It helps all types of organizations to improve profits and reduce the loss by taking timely decisions. Agriculture is also like an organization where farmers suffer with the loss most of the time in this business. This could be mainly because, there is no system to ensure the synchronization between the demand and supply for various food commodities required by the society. Science, enormous amount of data from different authorized sources like government websites revealing the demand and supply of various food commodities for forgoing period, this paper proposes a novel regression based AGRO forecasting model. This could help the farmers to make timely decisions and work towards fulfilling the actual needs of the society and avoiding putting themselves into the loss by growing unnecessary crops. Proposed model has been implemented using MapReduce parallel programming approach with Hadoop distributed File System. This processes time series data with Regression model for predicting the demand, supply and price for the agricultural commodities in distributed environment. Resulting forecasted values are in the range of real values.

关键词： Prediction Decision Agriculture Demand-supply Forecast parallel programming model Hadoop distributed file system Regression MapReduce time series

来源：评论

学校读者我要写书评

暂无评论

RedSync: Reducing synchronization bandwidth for distributed deep learning training system

引用

JOURNAL OF parallel AND distributed COMPUTING 2019年 133卷 30-39页

作者： Fang, Jiarui Fu, Haohuan Yang, Guangwen Hsieh, Cho-Jui Tsinghua Univ Dept Comp Sci Beijing Peoples R China Tsinghua Univ Dept Earth Syst Sci Beijing Peoples R China Tsinghua Univ Dept Comp Sci & Technol Beijing Peoples R China Natl Supercomp Ctr Wuxi Wuxi Jiangsu Peoples R China Univ Calif Los Angeles Comp Sci Los Angeles CA USA

Data parallelism has become a dominant method to scale Deep Neural Network (DNN) training across multiple nodes. Since the bandwidth requirement of synchronizing the gradients of the local model can be a bottleneck for large-scale distributed training, compressing communication traffic has gained widespread attention recently. Among several recent proposed compression algorithms, Residual Gradient Compression (RGC) is one of the most successful approaches-it can significantly compress the transmitting message size (0.1% of the gradient size) of each node and still achieve correct accuracy and the same convergence speed. However, the literature on compressing deep networks focuses almost exclusively on achieving good theoretical compression rate, while the efficiency of RGC in real implementation has been less investigated. In this paper, we develop an RGC method that is able to reduce the end-to-end training time on real-world multi-GPU systems. Our proposed RGC system design called RedSync, introduces a set of optimizations to reduce communication bandwidth while introducing limited overhead. We examine the performance of RedSync on two different multiple GPU platforms, including 128 GPUs of a supercomputer and an 8-GPU server. Our test cases include image classification on Cifar10 and ImageNet, and language modeling tasks on Penn Treebank and Wiki2 datasets. For DNNs featured with high communication to computation ratio, which has long been considered with poor scalability, RedSync shows significant performance improvement. (C) 2019 Elsevier Inc. All rights reserved.

关键词： Deep learning Data parallelism Graphics processing unit (GPU) Gradient compression Sparsification Quantization MPI

来源：评论

学校读者我要写书评

暂无评论

workshop heterogeneous equipment information perception analysis system 11

Workshop heterogeneous equipment information perception anal...

引用

11th International Conference on Intelligent Human-Machine systems and Cybernetics (IHMSC)

作者： Wu, Pengfei Qi, Mengjia Gao, Lingyan Zou, Wei Miao, Qiang Liu, Li-lan Shanghai Univ Shanghai Key Lab Intelligent Mfg & Robot Shanghai Peoples R China Shanghai Aerosp Syst Engn Inst Shanghai Peoples R China

ISBN: (纸本)9781728118598

In order to meet the high requirements of the shop management system for device data-aware interaction, in the traditional data-aware mode, this paper proposes a development platform based on Kepware as a perceptual software, through the combination of heterogeneous device network design and wireless network (AP), using OPC The interface technology realizes the parallel acquisition of multi-channel data of heterogeneous devices, and verifies the feasibility of the data acquisition method through experiments. The results of the research show that the information of real-time perception and production status monitoring based on Kepware can quickly process, analyze, real-time display and store the data of each channel. This scheme has certain reference value for the data collection of heterogeneous equipment in digital workshop.

关键词： multi-channel data acquisition heterogeneous network Kepware software

来源：评论

学校读者我要写书评

暂无评论

SFDDM: A Secure distributed Database Management in Combined Fog-to-Cloud systems 24

SFDDM: A Secure Distributed Database Management in Combined ...

引用

24th IEEE International workshop on Computer Aided Modeling and Design of Communication Links and Networks (IEEE CAMAD)

作者： Sengupta, Souvik Kahvazadeh, Sarang Masip-Bruin, Xavi Garcia, Jordi UPC BarcelonaTech CRAAX Lab Dept Comp Architecture DAC Vilanova I La Geltru Spain

ISBN: (纸本)9781728110165

Technological revolutions have greatly increased the use of IoT devices for our daily life. Driving the fact that everything surrounding us is getting connected what turns into an unstoppable increase in the amount of data produced. This data represents the state of diverse environmental events and helps to control a large set of distinct activities. So, accurate and secure management of this data is essential for any computing platform. Moreover, in order to provide real-time services in a distributed system (i.e., smart city), the data should be properly and securely managed. It is well known that shifting these tasks to the edge (i.e., near to the end users), highly facilitates these two objectives. The recently proposed Fog-to-Cloud (F2C) model is intended to enable data processing near to the edge, which helps to get better latency-sensitive services. However, some challenges remain to accurately and securely manage this data over the system, mainly due to the distributed F2C nature. Thus, considering these facts and challenges, in this paper we propose an architectural solution aimed at building a secure distributed database for F2C systems. Then, considering a real-case scenario, we perform some tests to measure the performance of our proposing schema. Finally, by comparing the performance between traditional cloud, fog/edge based execution model and our proposing SFDDM, we validate the effectiveness of our proposing schema.

关键词： Fog-to-Cloud (F2C) distributed Database Internet-of-Thing (IoT) Security and Privacy

来源：评论

学校读者我要写书评

暂无评论

Large-scale stochastic topology optimization using adaptive mesh refinement and coarsening through a two-level parallelization scheme

引用

COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING 2019年 343卷 186-206页

作者： Baiges, Joan Martinez-Frutos, Jesus Herrero-Perez, David Otero, Fermin Ferrer, Alex Univ Politecn Cataluna Jordi Girona 1-3Edif C1 ES-08034 Barcelona Spain Tech Univ Cartagena Computat Mech & Sci Comp Grp Campus Muralla del Mar Murcia 30202 Spain INEGI Rua Dr Roberto Frias 400 P-4200465 Porto Portugal CIMNE Edif C1Campus Nord UPC C Gran Capita S-N Barcelona 08034 Spain

Topology optimization under uncertainty of large-scale continuum structures is a computational challenge due to the combination of large finite element models and uncertainty propagation methods. The former aims to address the ever-increasing complexity of more and more realistic models, whereas the latter is required to estimate the statistical metrics of the formulation. In this work, the computational burden of the problem is addressed using a sparse grid stochastic collocation method, to calculate the statistical metrics of the topology optimization under uncertainty formulation, and a parallel adaptive mesh refinement method, to efficiently solve each of the stochastic collocation nodes. A two-level parallel processing scheme (TOUU-PS2) is proposed to profit from parallel computation on distributed memory systems: the stochastic nodes are distributed through the distributed memory system, and the efficient computation of each stochastic node is performed partitioning the problem using a domain decomposition strategy and solving each subdomain using an adaptive mesh refinement method. A dynamic load-balancing strategy is used to balance the workload between subdomains, and thus increasing the parallel performance by reducing processor idle time. The topology optimization problem is addressed using the topological derivative concept in combination with a level-set method. The performance and scalability of the proposed methodology are evaluated using several numerical benchmarks and real-world applications, showing good performance and scalability up to thousands of processors. (C) 2018 Elsevier B.V. All rights reserved.

关键词： Adaptive mesh refinement Large scale parallel computing Robust topology optimization Topological derivative Sparse grid

来源：评论

学校读者我要写书评

暂无评论

A distributed and parallel unite and conquer method to solve sequences of non-Hermitian linear systems

引用

JAPAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS 2019年第2期36卷 663-684页

作者： Wu, Xinzhe Petiton, Serge G. Maison Simulat F-91191 Gif Sur Yvette France Univ Lille CRIStAL F-59650 Villeneuve Dascq France

Many problems in science and engineering often require to solve a long sequence of large-scale non-Hermitian linear systems with different right-hand sides (RHSs) but a unique operator. Efficiently solving such problems on extreme-scale platforms requires the minimization of global communications, reduction of synchronization and promotion of asynchronous communications. Unite and Conquer GMRES/LS-ERAM (UCGLE) method (Wu and Petiton, in Proceedings of the International Conference on High Performance Computing in Asia-Pacific Region. ACM, New York, pp 36-46, 10.1145/3149457.3154481, 2018) is a suitable candidate with the reduction of global communications and the synchronization points of all computing units. It consists of three computing algorithms with asynchronous communications that allow the use of approximated eigenvalues to accelerate the convergence of solving linear systems and to improve fault tolerance. In this paper, we extend both the mathematical model and the implementation of UCGLE method to adapt to solve sequences of linear systems. The eigenvalues obtained in solving previous linear systems by UCGLE can be recycled, improved on the fly and applied to construct a new initial guess vector for subsequent linear systems, which can achieve a continuous acceleration to solve linear systems in sequence. Numerical experiments using different test matrices to solve sequences of linear systems on supercomputer Tianhe-2 indicate a substantial decrease in both computation time and iteration steps when the approximated eigenvalues are recycled to generate the initial guess vectors.

关键词： Sequence of linear systems Linear solvers Krylov methods Unite and Conquer Iterative methods Eigenvalues

来源：评论

学校读者我要写书评

暂无评论

31st International workshop on Languages and Compilers for parallel Computing, LCPC 2018

31st International Workshop on Languages and Compilers for P...

引用

31st International workshop on Languages and Compilers for parallel Computing, LCPC 2018

ISBN: (纸本)9783030346263

The proceedings contain 14 papers. The special focus in this conference is on Languages and Compilers for parallel Computing. The topics include: GASNet-EX: A high-performance, portable communication library for exascale;nested parallelism with algorithmic skeletons;HDArray: parallel array interface for distributed heterogeneous devices;automating the exchangeability of shared data abstractions;design and performance analysis of real-time dynamic streaming applications;a similarity measure for gpu kernel subgraph matching;new opportunities for compilers in computer security;footmark: A new formulation for working set statistics;towards an achievable performance for the loop nests;extending index-array properties for data dependence analysis;optimized sound and complete data race detection in structured parallel programs;compiler optimizations for parallel programs.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Power Control of distributed Energy Storage System in Bipolar DC Microgrid 22

Power Control of Distributed Energy Storage System in Bipola...

引用

22nd International Conference on Electrical Machines and systems (ICEMS)

作者： Zhang, Chunjiang Jiao, Meng Liang, Mengna Liu, Ming Kan, Zhizhong Yanshan Univ Sch Eletr Engn Qinhuangdao Hebei Peoples R China

ISBN: (纸本)9781728133980

distributed energy storage system plays an important role in stabilizing energy fluctuation and maintaining power balance in bipolar DC microgrid. The energy storage system is used to stabilize the bipolar DC bus voltage and ensure that the positive and negative bus voltages are equal. For achieving the above control objectives, the control strategy of single energy storage converter is studied, and the regulation ability of the strategy is analyzed. However, the single energy storage unit (ESU) is usually difficult to meet the large -capacity requirements of the microgrid system., so multiple ESUs can be connected in parallel to DC bus to form a distributed energy storage system for the expansion of system capacity. In this paper, an improved state of charge (SOC) droop control strategy is proposed. By constructing a new exponential function relationship between droop coefficient and SOC of ESU, the output or input power of each ESU can be adjusted in real time according to its own SOC during charging or discharging. Therefore, the balance of power and SOCs can be achieved among parallel ESUs, and the reasonable distribution of load power can be realized. Finally, the feasibility of the proposed control strategy is verified by experiments.

关键词： Bipolar DC microgrid distributed energy storage system Power balance control Droop control

来源：评论

学校读者我要写书评

暂无评论

A fine-grained controlled set-up to execute time synchronization protocol tests in a distributed environment 2

A fine-grained controlled set-up to execute time synchroniza...

引用

2nd IEEE International workshop on Metrology for Industry 4.0 and Internet of Things (IoT) (MetroInd4.0 and IoT)

作者： Cerro, G. Ferrigno, L. Miele, G. Capriglione, D. Lamonaca, F. Carni, D. L. Univ Cassino & Southern Lazio Dept Elect & Informat Engn Cassino Italy Univ Salerno Dept Ind Engn Fisciano Italy Univ Sannio Dept Engn Benevento Italy Univ Calabria Dept DIMES Arcavacata Di Rende Italy

ISBN: (纸本)9781728104294

time synchronization protocols assume a primary role in IoT era. Their design, implementation and test often follow separate paths and they are not ensured with the same level of accuracy. Simulation validation exhibits unpredictable gap with real performance levels. According to these statements and in order to answer to the proposed open issues, the paper present a fine-grained controlled environment to perform custom validation test of time synchronization algorithms. The set-up is built by adopting widely known Raspberry development boards, exploiting the ad-hoc fiber optic links provided by Unicas-Net, taking advantage of NetEm built-in emulator in Raspbian Operating systems, and finally developing a tailored user-friendly software to adequately control network link quality. Accordingly, synchronizing nodes can connect to one of the open points (provided by Raspberry devices) of the network ring and run their algorithms with other nodes that access to the set-up at 50 km distance. Finally, time synchronization performance can easily be assessed by exploiting GPS time provided by the GPS receiver and antenna, which complete each board equipment. First functionality tests are also reported in the paper.

关键词： time Synchronization Network Emulation distributed systems ad-hoc networks Measurement Set-Up

来源：评论

学校读者我要写书评

暂无评论

Multiple Balance Subsets Stacking for Imbalanced Healthcare Datasets

Multiple Balance Subsets Stacking for Imbalanced Healthcare ...

引用

International Conference on parallel and distributed systems (ICPADS)

作者： Yachao Shao Tao Zhao Xiaoning Wang Xiaofeng Zou Xiaoming Fu Institute of Computer Science University of Göttingen Göttingen Germany National University of Defense Technology Changsha Hunan China First Affiliated Hospital of Gannan Medical University Ganzhou Jiangxi China

ISBN: (数字)9781728190747

ISBN: (纸本)9781728183824

Accurate prediction is highly important for clinical decision making and early treatment. In this paper, we study the imbalanced data problem in prediction, a key challenge existing in the healthcare area. Imbalanced datasets bias classifiers towards the majority class, leading to an unsatisfied classification prediction performance on the minority class, which is known as imbalance problem. Existing imbalance learning methods may suffer from issues like information loss, overfitting, and high training time cost. To tackle these issues, we propose a novel ensemble learning method called Multiple bAlance Subsets Stacking (MASS) by exploiting a multiple balance subsets construction strategy. Furthermore, we improve MASS with introducing parallelism (parallel MASS) to reduce the training time cost. We evaluate MASS on three real-world healthcare datasets, and experimental results demonstrate that its prediction performance outperforms the state-of-art methods in terms of AUC, F1-score and MCC. Through the speedup analysis, parallel MASS reduces the training time cost greatly on large dataset, and its speedup increases as the data size grows.

关键词： Training Learning systems Conferences Stacking Decision making Medical services parallel processing

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：