检索结果-内蒙古大学图书馆

Split-bucket partition (SBP): a novel execution model for top-K and selection algorithms on GPUs

JOURNAL OF SUPERCOMPUTING 2024年第11期80卷 15122-15160页

作者： Yang, Yiqing Zhang, Guoyin Wu, Yanxia Zhao, Zhixiang Fu, Yan Harbin Engn Univ Dept Comp Sci Harbin 150001 Heilongjiang Peoples R China

Top-K and selection operations are critical in data processing and analysis, and their efficient implementation on GPUs is increasingly important due to the growing demands of data analysis. Existing methods, primarily relying on the bucket partition execution model, encounter challenges such as uneven bucket distribution and latency in merging processes. To address these issues, we introduce a novel Split-Bucket Partition (SBP) execution model that specifically addresses these challenges. Additionally, we propose task and control flow optimizations targeted at top-K and selection algorithms, which further contribute to performance improvements. Our optimized algorithms significantly outperform existing approaches, delivering performance gains of up to 2.3 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$2.3$$\end{document} times and 1.6 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$1.6$$\end{document} times for different bucket partitioning rules. Our algorithms show robust performance improvements in non-uniform data scenarios, with gains ranging from 1.9 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$1.9$$\end{document} times to 15.5 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$15.5$$\end{document} times. However, it should be noted that the SB

关键词： Bucket partition Top-K algorithms selection algorithms GPU parallel computing

来源：评论

学校读者我要写书评

暂无评论

A simple and fast linear-time algorithm for divisor methods of apportionment

引用

MATHEMATICAL PROGRAMMING 2024年第1-2期203卷 187-205页

作者： Reitzig, Raphael Wild, Sebastian Univ Liverpool Dept Comp Sci Liverpool England

Proportional apportionment is the problem of assigning seats to states (resp. parties) according to their relative share of the population (resp. votes), a field heavily influenced by the early work of Michel Balinski, not least his influential 1982 book with Peyton Young (Fair representation, 2nd edn. Brookings Institution Press, Washington, D.C., 2001). In this article, we consider the computational cost of divisor methods (also known as highest averages methods), the de-facto standard solution that is used in many countries. We show that a simple linear-time algorithm can exactly simulate all instances of the family of divisor methods of apportionment by reducing the problem to a single call to a selection algorithm. All previously published solutions were iterative methods that either offer no linear-time guarantee in the worst case or require a complex update step that suffers from numerical instability.

关键词： Proportional apportionment selection algorithms Divisor methods d'Hondt method Fair division Rounding percentages

来源：评论

学校读者我要写书评

暂无评论

A Method for Selecting Computing Power Nodes in a Computing Power Network Scenario

A Method for Selecting Computing Power Nodes in a Computing ...

引用

IEEE/CIC International Conference on Communications in China (ICCC)

作者： Zhao, Qianying Lei, Bo Wei, Min China Telecom Res Inst Beijing Peoples R China

ISBN: (纸本)9798350378412

The computing power network distributes computing power information of all network computing power nodes through network control planes or distributed routing protocols and combines network information such as latency to provide customers with optimized computing power allocation and network connection solutions. It addresses the issue of computing power scheduling among multi-level resource pools, achieving global optimization of resource allocation. However, after the computing power network collects information on computing power pools and network metrics, and receives user requests, the next research goal is how to select suitable resource pools for users based on known information to achieve global optimization of resource allocation. This paper proposes a method for selecting computing power pools. By utilizing computing power information collected by the computing power network controller from various nodes and considering business requirements, it selects appropriate resource pools for customers, avoiding resource pool fragmentation and maximizing resource pool utilization.

关键词： Computing power network Worst-Fit Best-fit selection algorithms

来源：评论

学校读者我要写书评

暂无评论

Time-lapse technology for embryo culture and selection

引用

UPSALA JOURNAL OF MEDICAL SCIENCES 2020年第2期125卷 77-84页

作者： Lundin, Kersti Park, Hannah Sahlgrens Univ Hosp Reprod Med Gothenburg Sweden

Culturing of human embryos in optimal conditions is crucial for a successful in vitro fertilisation (IVF) programme. In addition, the capacity to assess and rank embryos correctly for quality will allow for transfer of the potentially 'best' embryo first, thereby shortening the time to pregnancy, although not improving cumulative pregnancy and live birth rates. It will also encourage and facilitate the implementation of single embryo transfers, thereby increasing safety for mother and offspring. Time-lapse technology introduces the concept of stable culture conditions, in connection with the possibility of continuous viewing and documenting of the embryo throughout development. However, so far, even when embryo quality scoring is based on large datasets, or when using the time-lapse technology, the morphokinetic scores are still mainly based on subjective and intermittent annotations of morphology and timings. Also, the construction of powerful algorithms for widespread use is hampered by large variations in culture conditions between individual IVF laboratories. New methodology, involving machine learning, where every image from the time-lapse documentation is analysed by a computer programme, looking for patterns that link to outcome, may in the future provide a more accurate and non-biased embryo selection.

关键词： Assisted reproduction blastocyst transfer embryo quality selection algorithms

来源：评论

学校读者我要写书评

暂无评论

Fast Scalable selection algorithms for Large Scale Data

Fast Scalable Selection Algorithms for Large Scale Data

引用

IEEE International Conference on Big Data (Big Data)

作者： Thompson, Lee Parnell Xu, Weijia Miranker, Daniel P. Univ Texas Austin Dept Comp Sci Austin TX 78712 USA Univ Texas Austin Texas Adv Comp Ctr Austin TX 78712 USA

ISBN: (纸本)9781479912926;9781479912933

selection finding, and its most common form median finding, are used as a measure of central tendency for problems in biology, databases, and graphics. These problems often require selection finding as a subcomponent where it can be called many times, and as such speed is important. The Map/Reduce framework has been shown to be an important tool for creating scalable applications. There are a number of valid implementations of the selection algorithms inside of a Map/Reduce framework, certain of which are compared in this paper. However, as the volume of data increases, subtle theoretical algorithmic implementation differences can lead to significant differences in practical application. Therefore, an efficient and scalable selection finding method has the potential to provide general benefit to a number of applications. This paper compares algorithms that have been redesigned or created for the Map/Reduce framework for the purpose of selection finding, or, finding the k-th ranked element in an unordered set. This paper takes the concepts used from two existing selection algorithms and translates them into a novel method using the Map/Reduce framework with two variations. Each approach uses a different methodology to reduce the total amount of workload needed for a selection. All the algorithms are compared together for scalability and efficiency in a computing cluster environment with up to 256 processing cores. The results show that the methods proposed in this paper outperform several common alternatives in identifying medians with Hadoop, including using sorting, Pig, and BinMedian methods. Our implementations are also available upon request.

关键词： Hadoop Map Reduce selection algorithms Median Finding

来源：评论

学校读者我要写书评

暂无评论

Performance comparison of selection diversity receiver when different decision algorithms are used 26

Performance comparison of selection diversity receiver when ...

引用

26th Telecommunications Forum (TELFOR)

作者： Golubovic, Aleksandra Milic, Dejan Univ Nis Fac Elect Engn Aleksandra Medvedeva 14 Nish 18000 Serbia

ISBN: (纸本)9781538671719

This paper gives the comparison of dual-branch selection diversity receiver performance in interference-limited environment for four different decision algorithms: desired signal algorithm, signal-to-interference ratio algorithm, total signal algorithm and minimal interference algorithm, when Weibull fading model is used to describe a desired signal and interference.

关键词： interference selection algorithms selection combining diversity systems wireless telecommunications

来源：评论

学校读者我要写书评

暂无评论

Near Optimal Service Function Path Instantiation in a Multi-Datacenter Environment 11

Near Optimal Service Function Path Instantiation in a Multi-...

引用

11th International Conference on Network and Service Management (CNSM)

作者： Medhat, Ahmed M. Carella, Giuseppe Lueck, Christian Corici, Marius-Iulian Magedanz, Thomas Tech Univ Berlin Dept Telecommun Syst Berlin Germany Next Generat Network Infrastruct Fraunhofer Inst Berlin Germany

ISBN: (纸本)9783901882777

Service Function Chaining (SFC) supports services/applications through linking an ordered list of service functions (such as firewalls, deep packet inspections and load balancers) in the network. Using Software Defined Networking (SDN) and Network Function Virtualization (NFV) provides a simpler and shorter SFC process. In order to provide high-availability on a multi-site SFC, it is required a flexible algorithm able to properly select Service Function instances while creating the Service Function Path (SFP). This paper presents a new service function selection algorithm providing a flexible tradeoff between the SFP distance and the loads per service function instances at the deployment stage. The flexible tradeoff provided in the algorithm can be adjusted by a parameter. This parameter value can differ among different types of SFCs based on the service/application (real time or non-real time) dedicated for its SFC. Matlab has been utilized for validating it with different scenarios. It is also compared to other existing algorithms such as, load balancing and shortest path. The results showed that the proposed algorithm could provide the best SFP end-to-end distance performance as the shortest path algorithm for the real time services/applications, and that will result a good Quality of Service (QoS) performance for real time services/applications. Furthermore, the proposed algorithm provides an acceptable load distribution over the available service functions instances better than shortest path algorithm and not far from and load balancing algorithms.

关键词： Service Function Chaining SDN NFV selection algorithms

来源：评论

学校读者我要写书评

暂无评论

Energy Efficiency of Large-Scale Multiple Antenna Systems with Transmit Antenna selection

引用

IEEE TRANSACTIONS ON COMMUNICATIONS 2014年第2期62卷 638-647页

作者： Li, Hui Song, Lingyang Debbah, Merouane Peking Univ State Key Lab Adv Opt Commun Syst & Networks Sch Elect Engn & Comp Sci Beijing 100871 Peoples R China SUPELEC Gif Sur Yvette France

In this paper, we perform transmit antenna selection to improve the energy efficiency of large scale multiple antenna systems. We derive a good approximation of the distribution of the mutual information in this antenna selection system. It shows that channel hardening phenomenon is still retained as full complexity with antenna selection. Then, we use this closed-form expression to assess the energy efficiency performance. Specifically, we evaluate the performance of the energy efficiency in two different cases: 1) the circuit power consumption is comparable to or even dominates the transmit power, and 2) the circuit power can be ignored due to relatively much higher transmit power. The theoretical analysis indicates that there exists an optimal number of selected antennas to maximize the energy efficiency in the first case, whereas in the second case, the energy efficiency is maximized when all the available antennas are used. Based on these conclusions, two simple but efficient antenna selection algorithms are proposed to obtain the maximum energy efficiency. All the analytical results are verified through computer simulations.

关键词： Large scale multiple antenna system transmit antenna selection channel hardening energy efficiency selection algorithms

来源：评论

学校读者我要写书评

暂无评论

A selection algorithm for prototyping activities

引用

INTERNATIONAL JOURNAL OF INTERACTIVE DESIGN AND MANUFACTURING - IJIDEM 2014年第1期8卷 1-11页

作者： Filippi, S. Barattin, D. Univ Udine DIEGM Dept PIRG Prod Innovat Res Grp Via Sci 206 I-33100 Udine Italy

Modern product development processes heavily exploit both classic prototyping technologies and procedures and new ones appearing on the market almost every day. Such heavy use has generated a wide collection of prototyping activities, making the selection of the best ones for specific development contexts increasingly difficult. The authors recently highlighted five dimensions for characterizing prototyping activities. In this study, these dimensions were to propose a new classification of prototyping activities. Within this context, the research described in this paperwas aimed at developing a selection algorithm that covers as wide a variety of design situations as possible. The algorithm has already been implemented in the field, and the initial experiences with the algorithm are also described herein.

关键词： Prototyping activities selection algorithms Design process

来源：评论

学校读者我要写书评

暂无评论

Selecting a discrete portfolio

引用

JOURNAL OF MATHEMATICAL ECONOMICS 2014年第1期55卷 69-73页

作者： Olszewski, Wojciech Vohra, Rakesh Northwestern Univ Dept Econ Evanston IL 60208 USA Univ Penn Dept Econ Philadelphia PA 19104 USA Univ Penn Dept Elect & Syst Engn Philadelphia PA 19104 USA

We study the problem of selecting an optimal portfolio out of a finite set of available assets. Assets are characterized by their expected returns and the covariance matrix, and investors are assumed to have a mean-variance utility, that is, their utility function is linear in the mean and variance of the portfolio they hold. When assets are negatively correlated, or even when a slightly more general condition is satisfied, we provide an algorithm for selecting an optimal portfolio. We illustrate the usefulness of this algorithm by some comparative statics result. When assets can be positively correlated, we deliver a negative result regarding the existence of useful algorithms for selecting an optimal portfolio. (C) 2014 Elsevier B.V. All rights reserved.

关键词： Optimal discrete portfolios selection algorithms

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：