检索结果-南通市图书馆

On Peer-Assisted Data Dissemination in Data Center Networks: Analysis and Implementation

在线全文

学校读者我要写书评

暂无评论

Tsinghua Science and Technology 2014年第1期19卷 51-64页

作者： Yaxiong Zhao Jie Wu Cong Liu google inc. Temple University Sun Yat-Sen University

Data Center Networks （DCNs） are the fundamental infrastructure for cloud computing. Driven by the massive parallel computing tasks in cloud computing, one-to-many data dissemination becomes one of the most important traffic patterns in DCNs. Many architectures and protocols are proposed to meet this demand. However, these proposals either require complicated configurations on switches and servers, or cannot deliver an optimal performance. In this paper, we propose the peer-assisted data dissemination for DCNs. This approach utilizes the rich physical connections with high bandwidths and mutli-path connections, to facilitate efficient one-to-many data dissemination. We prove that an optimal P2P data dissemination schedule exists for FatTree, a specially- designed DCN architecture. We then present a theoretical analysis of this algorithm in the general multi-rooted tree topology, a widely-used DCN architecture. Additionally, we explore the performance of an intuitive line structure for data dissemination. Our analysis and experimental results prove that this simple structure is able to produce a comparable performance to the optimal algorithm. Sinc. DCN applications heavily rely on virtualization to achieve optimal resource sharing, we present a general implementation method for the proposed algorithms, which aims to mitigate the impact of the potentially-high churn rate of the virtual machines.

关键词： data center networks cloud computing P2P scheduling peer-assisted data dissemination

Tiny Det: accurately detecting small objects within 1 GFLOPs

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2023年第1期66卷 318-319页

作者： Shaoyu CHEN Tianheng CHENG Jiemin FANG Qian ZHANG Yuan LI Wenyu LIU Xinggang WANG School of Electronic Information and Communications Huazhong University of Science and Technology Institute of Artificial Intelligence Huazhong University of Science and Technology Horizon Robotics inc. google inc.

Dear editor,Object detection plays an important role in computer vision,and has gradually become the technical foundation of many applications. However, the inference procedure of advanced detection models requires ma... 详细信息

关键词： Economic and social effects

同方期刊数据库

Datacenter optics:requirements,technologies,and trends(Invited Paper)

在线全文

学校读者我要写书评

暂无评论

Chinese Optics Letters 2017年第5期15卷 87-90页

作者： Xiang Zhou Hong Liu Ryohei Urata Platforms Google Inc. Mountain View CA 94043 USA

We review over a decade of technology evolution and advancement of intra-datacenter optical interconnect, mainly driven by the explosive bandwidth growth of web and cloud-based services. Emerging trends and tech-nolog... 详细信息

关键词： over will VCSEL Invited Paper Datacenter optics:requirements,technologies,and trends link BW

Efficient Execution of Multiple Queries on Deep Memory Hierarchy

在线全文

学校读者我要写书评

暂无评论

Journal of Computer Science & Technology 2007年第2期22卷 273-279页

作者：张岩陈智峰周源源 National Laboratory on Machine Perception Peking University Beijing 100871 China google inc. 1600 Amphitheatre Parkway Mountain View CA 94043 U.S.A. Department of Computer Science University of Illinois at Urbana-Champaign Urbana IL 61801 U.S.A.

This paper proposes a complementary novel idea, called MiniTasking to further reduce the number of cache misses by improving the data temporal locality for multiple concurrent queries. Our idea is based on the observation that, in many workloads such as decision support systems （DSS）, there is usually significant amount of data sharing among different concurrent queries. MiniTasking exploits such data sharing to improve data temporal locality by scheduling query execution at three levels： query level batching, operator level grouping and mini-task level scheduling. The experimental results with various types of concurrent TPC-H query workloads show that, with the traditional N-ary Storage Model （NSM） layout, MiniTasking significantly reduces the L2 cache misses by up to 83%, and thereby achieves 24% reduction in execution time. With the Partition Attributes Across （PAX） layout, MiniTasking further reduces the cache misses by 65% and the execution time by 9%. For the TPC-H throughput test workload, MiniTasking improves the end performance up to 20%.

关键词： cache performance temporal locality mini-task scheduling concurrent queries

Probiotic Medilac-S^(█) for the induction of clinical remission in a Chinese population with ulcerative colitis: A systematic review and meta-analysis

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

World Journal of Clinical Cases 2018年第15期6卷 961-984页

作者： Ghania Sohail Xiaoyu Xu Mary C Christman Thomas A Tompkins Lallemand Health Solutions inc. MCC Statistical Consulting

AIM To assess the effects of probiotic Medilac-S~ as adjunctive therapy for the induction of remission of ulcerative colitis(UC) in a Chinese population through a systematic review and meta-analysis. METHODS A systematic literature search was conducted to find randomized, controlled trials in a Chinese population with at least two study arms-a control arm which receives a conventional, oral aminosalicylate drug, and a treatment arm, which administers the same conventional drug in conjunction with the probiotic Medilac-S~ per os. Both English and Chinese databases were searched, inc.uding Pub Med, EMBASE, google Scholar, Chinese National Knowledge Infrastructure, Wanfang Data, and VIP Search, and study data was extracted onto standardized abstraction sheets. Meta-analyses were conducted for primary and secondary outcomes of interest using a fixed or random effects model. The primary outcome was the induction of clinical remission and the secondary outcomes inc.uded changes in Sutherland index, endoscopic and histological scores, proportion of reported clinical symptoms and adverse events(AEs). For outcomes with sufficient data, the type of conventional drug therapy was also assessed to determine if the effects of combination therapy with Medilac-S~ was influenced by drug type. All tests were conducted using a type Ⅰ error rate of 0.05 and all confidence intervals(CI) were based on a 95% confidence level. Review protocol was uploaded to PROSPERO(CRD42018085658 upon completion).RESULTS Fifty-three clinical trials with a total of 3984 participants were identified and inc.uded in the review. Medilac-S~ adjunctive therapy significantly improved induction of clinical remission(RR = 1.21; 95%CI: 1.18-1.24; P < 0.0001) with the estimated likelihood of effective treatment, on average, 21% higher for those consuming the probiotic. Sutherland index scores showed the control mean was on average 3.10(CI: 2.41-3.78; P = 0.0428) units greater than the treatment mean, thereby de

关键词： Clinical remission Systematic review Meta-analysis Mesalazine Sulfasalazine Ulcerative colitis Medilac-S^(█)

A Study of Animated Transition in Similarity-Based Tiled Image Layout

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Tsinghua Science and Technology 2013年第2期18卷 157-170页

作者： Huan Zhang Jun Tao Fang Ruan Chaoli Wang Dematic Corp Grand Rapids MI 49505USA the Department of Computer Science Michigan Technological University Houghton MI49931 USA google inc Mountain View CA 94043USA

For many information visualization applications, showing the transition when interacting with the data is critically important as it can help users better perceive the changes and understand the underlying data. In this paper, we investigate the effectiveness of animated transition in a tiled image layout where the spiral arrangement of the images is based on their similarity. Three aspects of animated transition are considered, inc.uding animation steps, animation actions, and flying paths. Exploring and weighting the advantages and disadvantages of different methods for each aspect and in conjunction with the characteristics of the spiral image layout, we present an integrated solution, called AniMap, for animating the transition from an old layout to a new layout when a different image is selected as the query image. We show the effectiveness of our animated transition solution by demonstrating experimental results and conducting a comparative user study.

关键词： animated transition image layout large image collections user study

Deep Learning Based Target Tracking and Classification for Infrared Videos Using Compressive Measurements

在线全文

学校读者我要写书评

暂无评论

Journal of Signal and Information Processing 2019年第4期10卷 167-199页

作者： Chiman Kwan Bryan Chou Jonathan Yang Trac Tran Applied Research LLC Rockville MD USA google Inc. Mountain View CA USA Department of Electrical and Computer Engineering The Johns Hopkins University Baltimore MD USA

Although compressive measurements save data storage and bandwidth usage, they are difficult to be used directly for target tracking and classification without pixel reconstruction. This is because the Gaussian random matrix destroys the target location information in the original video frames. This paper summarizes our research effort on target tracking and classification directly in the compressive measurement domain. We focus on one particular type of compressive measurement using pixel subsampling. That is, original pixels in video frames are randomly subsampled. Even in such a special compressive sensing setting, conventional trackers do not work in a satisfactory manner. We propose a deep learning approach that integrates YOLO (You Only Look Once) and ResNet (residual network) for multiple target tracking and classification. YOLO is used for multiple target tracking and ResNet is for target classification. Extensive experiments using short wave infrared (SWIR), mid-wave infrared (MWIR), and long-wave infrared (LWIR) videos demonstrated the efficacy of the proposed approach even though the training data are very scarce.

关键词： Target Tracking Classification Compressive Sensing SWIR MWIR LWIR YOLO ResNet Infrared Videos

Measuring and Visualizing Research Collaboration and Productivity

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Data and Information Science 2018年第1期3卷 54-74页

作者： Jon Garner Alan L.Porter Andreas Leidolf Michelle Baker Search Technology Inc. School of Public Policy Georgia Institute of Technology Ecology Center Utah State University Department of Biology Utah State University

Purpose： This paper presents findings of a quasi-experimental assessment to gauge the research productivity and degree of interdisciplinarity of research center outputs. Of special interest, we share an enriched visualization of research co-authoring patterns. Design/methodology/approach： We compile publications by 45 researchers in each of 1） the iUTAH project, which we consider here to be analogous to a ＂research center,＂ 2） CG1-a comparison group of participants in two other Utah environmental research centers, and 3） CG2--a comparison group of Utah university environmental researchers not associated with a research center. We draw bibliometric data from Web of Science and from google Scholar. We gather publications for a period before iUTAH had been established （2010-2012） and a period after （2014-2016）. We compare these research outputs in terms of publications and citations thereto. We also measure interdisciplinarity using Integration scoring and generate science overlay maps to locate the research publications across disciplines. Findings： We find that participation in the iUTAH project appears to inc.ease research outputs （publications in the After period） and inc.ease research citation rates relative to the comparison group researchers （although CG 1 research remains most cited, as it was in the Before period）. Most notably, participation in iUTAH markedly inc.eases co-authoring among researchers--in general; and for junior, as well as senior, faculty; for men and women： across organizations; and across disciplines. Research limitations： The quasi-experimental design necessarily generates suggestive, not definitively causal, findings because of the imperfect controls. Practical implications： This study demonstrates a viable approach for research assessment of a center or program for which random assignment of control groups is not possible. It illustrates use of bibliometric indicators to inform R＆D program management. Originality/value： New visualizations

关键词： Bibliometrics Interdisciplinarity Multidisciplinarity Research evaluation Research collaboration mapping Science visualization Scientometrics Social networkanalysis

Identifying viruses from metagenomic data using deep learning

在线全文

学校读者我要写书评

暂无评论

Quantitative Biology 2020年第1期8卷 64-77页

作者： Jie Ren Kai Song Chao Deng Nathan AAhlgren Jed AFuhrman Yi Li Xiaohui Xie Ryan Poplin Fengzhu Sun Quantitative and Computational Biology Program University of Southern CaliforniaLos AngelesCA 90089USA School of Mathematics and Statistics Qingdao UniversityQingdao 266071China Department of Biology Clark UniversityWorcesterMA 01610USA Department of Biological Sciences University of Southern CaliforniaLos AngelesCA 90089USA Department of Computer Science University of CaliforniaIrvineCA 92697USA google inc. Mountain ViewCA 94043USA

Background:The recent development of metagenomic sequencing makes it possible to massively sequence microbial genomes inc.uding viral genomes without the need for laboratory *** reference-based and gene homology-based methods are not efficient in identifying unknown viruses or short viral sequences from metagenomic ***:Here we developed a reference-free and alignment-free machine learning method,DeepVirFinder,for identifying viral sequences in metagenomic data using deep ***'.Trained based on sequences from viral RefSeq discovered before May 2015,and evaluated on those discovered after that date,DeepVirFinder outperformed the state-of-the-art method VirFinder at all contig lengths,achieving AUROC 0.93,0.95,0.97,and 0.98 for 300,500,1000,and 3000 bp sequences *** the training data with additional millions of purified viral sequences from metavirome samples further improved the accuracy for identifying virus groups that are *** DeepVirFinder to real human gut metagenomic samples,we identified 51,138 viral sequences belonging to 175 bins in patients with colorectal carcinoma(CRC).Ten bins were found associated with the cancer status,suggesting viruses may play important roles in ***:Powered by deep learning and high throughput sequencing metagenomic data,DeepVirFinder significantly improved the accuracy of viral identification and will assist the study of viruses in the era of metagenomics.

关键词： metagenome deep learning virus identification machine learning