检索结果-内蒙古大学图书馆

Robust, Scalable, and Fast Bootstrap Method for Analyzing Large Scale Data

IEEE TRANSACTIONS ON SIGNAL PROCESSING 2016年第4期64卷 1007-1017页

作者： Basiri, Shahab Ollila, Esa Koivunen, Visa Aalto Univ Dept Signal Proc & Acoust FIN-00076 Espoo Finland

In this paper we address the problem of performing statistical inference for large scale data sets i.e., Big Data. The volume and dimensionality of the data may be so high that it cannot be processed or stored in a single computing node. We propose a scalable, statistically robust and computationally efficient bootstrap method, compatible with distributed processing and storage systems. Bootstrap resamples are constructed with smaller number of distinct data points on multiple disjoint subsets of data, similarly to the bag of little bootstrap method (BLB) [A. Kleiner, A. Talwalkar, P. Sarkar, and M. I. Jordan, "A scalable bootstrap for massive data," J. Roy. Statist. Soc.: Ser. B (Statist. Methodol.), vol. 76, no. 4, pp. 795-816, 2014]. The disjoint subsets are significantly smaller than the original full data set and they may be processed in different storage and computing units in parallel. Then significant savings in computation is achieved by avoiding the recomputation of the estimator for each bootstrap sample. Instead, a computationally efficient fixed-point estimation equation is analytically solved via a smart approximation following the Fast and Robust Bootstrap method (FRB) [M. Salibian-Barrera, S. Van Aelst, and G. Willems, "Fast and robust bootstrap," Statist. Methods Appl., vol. 17, no. 1, pp. 41-71, 2008]. Our proposed bootstrap method facilitates the use of highly robust statistical methods in analyzing large scale data sets. The favorable statistical properties of the method are established analytically. Numerical examples demonstrate scalability, low complexity and robust statistical performance of the method in analyzing large data sets.

关键词： Bag of little bootstraps big data bootstrap distributed computation fast and robust bootstrap robust estimation

来源：评论

学校读者我要写书评

暂无评论

Latent Space Inference of Internet-Scale Networks

引用

JOURNAL OF MACHINE LEARNING RESEARCH 2016年第1期17卷 2756-2796页

作者： Ho, Qirong Yin, Junming Xing, Eric P. ASTAR Inst Infocomm Res Singapore 138632 Singapore Univ Arizona Eller Coll Management Dept Management Informat Syst Tucson AZ 85721 USA Carnegie Mellon Univ Sch Comp Sci Pittsburgh PA 15213 USA

The rise of Internet-scale networks, such as web graphs and social media with hundreds of millions to billions of nodes, presents new scientific opportunities, such as overlapping community detection to discover the structure of the Internet, or to analyze trends in online social behavior. However, many existing probabilistic network models are difficult or impossible to deploy at these massive scales. We propose a scalable approach for modeling and inferring latent spaces in Internet-scale networks, with an eye towards overlapping community detection as a key application. By applying a succinct representation of networks as a bag of triangular motifs, developing a parsimonious statistical model, deriving an efficient stochastic variational inference algorithm, and implementing it as a distributed cluster program via the Petuum parameter server system, we demonstrate overlapping community detection on real networks with up to 100 million nodes and 1000 communities on 5 machines in under 40 hours. Compared to other state-of-the-art probabilistic network approaches, our method is several orders of magnitude faster, with competitive or improved accuracy at overlapping community detection.

关键词： probabilistic network models triangular modeling stochastic variational inference distributed computation big data

来源：评论

学校读者我要写书评

暂无评论

A parallel alternating direction method with application to compound l₁-regularized imaging inverse problems

引用

INFORMATION SCIENCES 2016年 348卷 179-197页

作者： He, Chuan Hu, Changhua Li, Xuelong Yang, Xiaogang Zhang, Wei High Tech Inst Xian Xian 710025 Shaanxi Peoples R China Chinese Acad Sci Xian Inst Opt & Precis Mech Xian 710199 Shaanxi Peoples R China

We derive a parallel alternating direction method of multipliers (PADMM) and apply it to compound l(1)-regularized imaging inverse problems. The proposed method is capable of locating the saddle point of large-scale convex minimization problems with the sum of several nonsmooth but proximable terms. Using an operator splitting strategy, the objective is decomposed into subproblems that are conveniently, individually and simultaneously solved. With the assistance of the Moreau decomposition, our method excludes auxiliary variables that exist in the ADMM and possesses a compacter structure. Thus, the proposed method is preferable in distributed computation. The convergence proof and convergence rate analysis are presented. Application to both image restoration and image compressed sensing demonstrates the effectiveness and efficiency of the proposed method. (C) 2016 Elsevier Inc. All rights reserved.

关键词： Parallel alternating direction method of multipliers (PADMM) Primal-dual splitting Image deblurring MRI reconstruction distributed computation Adaptive parameter estimation

来源：评论

学校读者我要写书评

暂无评论

A Kind of Load Balance Multicast Routing Algorithm Based on distributed computation and Centralized Control

A Kind of Load Balance Multicast Routing Algorithm Based on ...

引用

The International Conference on Recent Trends in Materials and Mechanical Engineering(ICRTMME 2011)

作者： CHEN Xiang Tang Jun-yong ZHANG Yong School of Civil Engineering Xi'an Technological University Department of Computer Science and Engineering Xi'an Technological University Wuhan University of Science and Technology

distributed routing can find multicast tree and adapt to dynamic changes of multicast members in case of acknowledge local network stateMajority of current distributed multicast routing algorithms based on the shortest delay, which can achieve load balance large additional overheadThe paper presented a kind of load balance multicast routing algorithm based on distributed computation and centralized controlIt reside computation Agent in each node to compute forward probability and to maintain node routing table, node state table and forwarding probability tableIn addition, it can detect delay of adjacent nodes and coordinate session establishment based on the copy, parameter transmission and notification of control AgentOn this basis, the algorithm achieved dynamically joins and leave session of multicast membersAlgorithm analysis results show that the algorithm can improve the overall efficiency of network transmission.

关键词： multicast routing Agent distributed computation centralized control

来源：评论

学校读者我要写书评

暂无评论

Emulating cellular automata in chemical reaction-diffusion networks

引用

NATURAL COMPUTING 2016年第2期15卷 197-214页

作者： Scalise, Dominic Schulman, Rebecca Johns Hopkins Univ Dept Chem & Biomol Engn Baltimore MD USA Johns Hopkins Univ Dept Comp Sci Baltimore MD 21218 USA

Chemical reactions and diffusion can produce a wide variety of static or transient spatial patterns in the concentrations of chemical species. Little is known, however, about what dynamical patterns of concentrations can be reliably programmed into such reaction-diffusion systems. Here we show that given simple, periodic inputs, chemical reactions and diffusion can reliably emulate the dynamics of a deterministic cellular automaton, and can therefore be programmed to produce a wide range of complex, discrete dynamics. We describe a modular reaction-diffusion program that orchestrates each of the fundamental operations of a cellular automaton: storage of cell state, communication between neighboring cells, and calculation of cells' subsequent states. Starting from a pattern that encodes an automaton's initial state, the concentration of a "state" species evolves in space and time according to the automaton's specified rules. To show that the reaction-diffusion program we describe produces the target dynamics, we simulate the reaction-diffusion network for two simple one-dimensional cellular automata using coupled partial differential equations. Reaction-diffusion based cellular automata could potentially be built in vitro using networks of DNA molecules that interact via branch migration processes and could in principle perform universal computation, storing their state as a pattern of molecular concentrations, or deliver spatiotemporal instructions encoded in concentrations to direct the behavior of intelligent materials.

关键词： Reaction-diffusion Cellular automata DNA strand displacement Chemical reaction network Intelligent materials Molecular programming Programmable matter distributed computation

来源：评论

学校读者我要写书评

暂无评论

A Lottery SMC Protocol for the Selection Function in Software Defined Wireless Sensor Networks

引用

IEEE SENSORS JOURNAL 2016年第20期16卷 7325-7331页

作者： Sun, Yi Lin, Zhaowen Ma, Yan Beijing Univ Posts & Telecommun Inst Network Technol Network & Informat Ctr Beijing 100876 Peoples R China Sci & Technol Informat Transmiss & Disseminat Com Shijiazhuang 050081 Peoples R China Natl Engn Lab Mobile Network Secur Beijing 100876 Peoples R China Beijing Univ Posts & Telecommun Inst Sensing Technol & Business Beijing 100876 Peoples R China

In this paper, we focus on selecting targeted number from private sensory data in the software-defined wireless sensor networks and propose a lottery secure multiparty computation protocol for the selection function based on the layered homomorphic encryption and the equivalent transformation of two arrays. We also prove that it is secure via real/ideal simulation paradigm, no matter the Web server is honest or semi-honest in the presented security model.

关键词： Information security distributed computation secure multiparty computation software defined wireless sensor network homomorphic encryption

来源：评论

学校读者我要写书评

暂无评论

A distributed Voltage Stability Control Method for Integrated Transmission and Distribution Grids

A Distributed Voltage Stability Control Method for Integrate...

引用

IEEE Conference on Energy Internet and Energy System Integration

作者： Jinquan Zhao Qing Lin Shiying Ma College of Energy and Electrical Engineering Hohai University China Electric Power Research Institute

ISBN: (纸本)9781538614280

While a large amount of distributed generators (DGs) were integrated into the distribution grid, the transmission grid voltage stability assessment (VSA) had not been suitable to be solved independently. A distributed voltage stability control (DVSC) method for integrated transmission and distribution (T&D) grids was proposed in this paper. The objective function aims to minimize the total control costs of the transmission and distribution grids. The boundary influence factors composed by the dual multipliers are introduced to coordinate the TN VSC sub-problem and DN VSC sub-problem. Those optimal power flow algorithms, by which provide the dual multipliers are generated, are adopted in both TN and DN VSC sub-problems. The equivalent power, the influence factors and the boundary bus voltages are exchanged between the transmission and distribution grids to realize distributed computation. Simulation results on integrated IEEE30 buses transmission and IEEE33 buses distribution system show that the proposed method is effective.

关键词： Integrated transmission and distribution grids Voltage stability control distributed generators distributed computation Optimal power flow

来源：评论

学校读者我要写书评

暂无评论

A Modified Ant Colony Based Optimization for Managing Cloud Resources in Manufacturing Sector 2

A Modified Ant Colony Based Optimization for Managing Cloud ...

引用

2nd International Conference on Green High Performance Computing (ICGHPC)

作者： Brintha, N. C. Jappes, J. T. Winowlin Benedict, Shajulin Kalasalingam Inst Technol Krishnankoil India Kalasalingam Univ Dept Mech Engn Krishnankoil India St Xaviers Catholic Coll Engn HPCCLoud Res Lab Chunkankadai India

ISBN: (纸本)9781467366151

Resource scheduling and management is an important problem in Cloud Manufacturing. The concept of optimization for scheduling jobs is an important issue to be considered in scheduling of different resources among heterogeneous users. The resources are placed across diversified locations in cloud and the major task is to distribute the resources effectively such that the makespan and completion time is reduced. In this paper, a Modified Ant Colony based optimization technique is proposed to optimize the resources through distributed computation. ACO is used to choose one among the different alternative rules to determine the processing order of each resource. Rather than having a larger search space, this approach reduces the search space and gives better solution. This reduces the delay in allocating resources to the user by providing an adaptive and global search technique. This approach reduces the total completion time of jobs and also takes in to account the migration time of the process. A series of experiments were conducted and the results of the experiment are compared with other heuristic algorithms like PSO. The results have shown that this approach can produce optimal solutions quickly by reducing delays.

关键词： Cloud Cloud manufacturing PSO Ant Colony Optimization distributed computation adaptive search heuristic

来源：评论

学校读者我要写书评

暂无评论

Research on distributed warning system of water quality in Mudan river based on EFDC and GIS

Research on distributed warning system of water quality in M...

引用

作者： Gula Tang Universite de Strasbourg

学位级别：博士

The consecutively happening water pollution events in China gave us the alarming that the water pollutions are disastrous, influence of the events is far- reaching with a heavy sacrifice in addition to the enormous economic loss. In order to strengthen capability of water pollution forecasting and early warning, relieve the harm to water ecology, animals and human, reduce economic loss and improve human settlement environmental quality, it is extremely necessary to carry out studies on Simulation and Early Warning System (SEWS) of water quality so that the spatio-temporal variation and its change trend can be accurately simulated, forecasted and evaluated for reasonable actions and policies to administrate the water environment in the country. Mudan River, an important river in northeastern cold regions of China, flows into Songhua River, which, after merging with Heilong River, continues its way running out of China into Russia. Thus, the quality of water in Mudan River is highly concerned not only locally and regionally but also internationally. Due to long winter season, SEWS of water quality in Mudan River would face many difficulties as a result of ice-covered surface for approximately 5 months in a year. As to the Mudan River, SEWS of water quality is a powerful tool for the river water quality control to guarantee safety of water environment. it is urgently to integrate the GIS techniques with the water quality simulation models to monitor sewage outlets discharge state of rivers especially those outlets with over-loaded drainage and un-reported discharge, to support day-to-day administration to the environment and water quality of the basin, to reinforce water environment monitoring and early warning, to reduce the occurrence of water pollution events, to strengthen capability of water quality forecasting and early warning when pollution occurs, to provide decision information support to administration of water pollution emergency, and finally to reduce the

关键词： early warning hydrodynamic model water quality two-dimensional simulation distributed computation scenario analysis GIS

来源：评论

学校读者我要写书评

暂无评论

Towards a distributed infrastructure for evolving graph analytics 25

Towards a distributed infrastructure for evolving graph anal...

引用

25th International Conference on World Wide Web (WWW)

作者： Moffitt, Vera Zaychik Stoyanovich, Julia Drexel Univ Philadelphia PA 19104 USA

ISBN: (纸本)9781450341448

Graphs are used to represent a plethora of phenomena, from the Web and social networks, to biological pathways, to semantic knowledge bases. Arguably the most interesting and important questions one can ask about graphs have to do with their evolution. Which Web pages are showing an increasing popularity trend? How does influence propagate in social networks? How does knowledge evolve? In this paper we present our ongoing work on the Portal system, an open-source distributed framework for evolving graphs. Portal streamlines exploratory analysis of evolving graphs, making it efficient and usable, and providing critical tools to computational and data scientists. Our system implements a declarative query language by the same name, which we briefly describe in this paper. Our basic abstraction is a TGraph, which logically represents a series of adjacent snapshots. We present different physical representations of TGraphs and show results of a preliminary experimental evaluation of these physical representations for an important class of evolving graph analytics.

关键词： evolving graphs graph analytics distributed computation

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：