检索结果-内蒙古大学图书馆

Mcc: a predictable and scalable massive client load generator 2nd

学校读者我要写书评

暂无评论

Mcc: a predictable and scalable massive client load generato...

2nd International Symposium on Benchmarking, Measuring, and Optimization, Bench 2019

作者： Wu, Wenqing Feng, Xiao Zhang, Wenli Chen, Mingyu State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Science Beijing China University of Chinese Academy of Sciences Beijing China Peng Cheng Laboratory Shenzhen China

ISBN: (纸本)9783030495558

The network load generators are widely used by network researchers to analyze link bandwidth, evaluate network performance and test device capabilities. Data center and IoT networks are quickly evolving and we desire to get a load generator that can precisely generate flow-level workload with high-throughput. Often researchers choose software-based generators because of their flexibility and open-source nature. However, despite the emerging of different solutions, existing software-based flow-level generators have difficulty in generating millions of concurrent TCP connections or achieving one-microsecond precision of packet inter departure time (IDT) which can undermine the correctness of experiments. In this paper, we present a new network load generator, called Massive Client Connections (MCC). MCC is a client load generator which means it performs flow-level load simulation. We separate the control plane from the data plane and design a two-stage timer mechanism to get higher precision. To take full advantage of multicore processors, we utilize the shared-nothing multi-threaded model. Our evaluation demonstrates that MCC generates network load conforming to expected distribution with one-microsecond precision. Moreover, MCC shows definite scalability of throughput in multicore systems. And it is capable of generating more than three million concurrent TCP connections with ten CPU cores. © Springer Nature Switzerland AG 2020.

关键词： Scalability

Weight Light, Hear Right: Heart Sound Classification with a Low-Complexity Model

学校读者我要写书评

暂无评论

Weight Light, Hear Right: Heart Sound Classification with a ...

European Signal Processing Conference (EUSIPCO)

作者： Jiahao Ji Lixian Zhu Haojie Zhang Kun Qian Kele Xu Zikai Song Bin Hu Björn W. Schuller Yoshiharu Yamamoto Key Laboratory of Brain Health Intelligent Evaluation and Intervention (Beijing Institute of Technology) Ministry of Education Beijing China School of Medical Technology Beijing Institute of Technology Beijing China School of Computer Science National University of Defense Technology P. R. China GLAM - the Group on Language Audio & Music Imperial College London UK CHI - Chair of Health Informatics. MRI Technical University of Munich Germany Educational Physiology Laboratory Graduate School of Education The University of Tokyo Japan

ISBN: (数字)9789464593617

ISBN: (纸本)9798331519773

Cardiovascular diseases (CVDs) remain a leading cause of mortality, necessitating early self-diagnosis for effective management. computer-aided CVD diagnosis via computerised auscultation has gained traction, advancing intelligent diagnostics. However, integrating complex neural networks into medical edge devices remains challenging. This paper introduces a novel lightweight model for heart sound classification based on broadcast residual learning, which can be seamlessly deployed in portable health monitoring devices for real-time cardiac auscultation. Comparative experiments validate the model's efficacy, achieving 89.1% accuracy and 89.7% Fl score with just 8.05 K parameters and 10.6 M MACC, showcasing superior performance within constrained complexity.

关键词： Heart Performance evaluation Microcontrollers Computational modeling Neural networks Europe Signal processing Real-time systems Monitoring Medical diagnostic imaging

Finding Backdoors to Integer Programs: A Monte Carlo Tree Search Framework

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Khalil, Elias B. Vaezipoor, Pashootan Dilkina, Bistra Department of Mechanical & Industrial Engineering University of Toronto Canada Scale AI Research Chair in Data-Driven Algorithms for Modern Supply Chains Department of Computer Science University of Toronto Canada Vector Institute for Artificial Intelligence Canada Department of Computer Science University of Southern California United States

In Mixed Integer Linear Programming (MIP), a (strong) backdoor is a "small" subset of an instance's integer variables with the following property: in a branch-and-bound procedure, the instance can be solved to global optimality by branching only on the variables in the backdoor. Constructing datasets of pre-computed backdoors for widely used MIP benchmark sets or particular problem families can enable new questions around novel structural properties of a MIP, or explain why a problem that is hard in theory can be solved efficiently in practice. Existing algorithms for finding backdoors rely on sampling candidate variable subsets in various ways, an approach which has demonstrated the existence of backdoors for some instances from MIPLIB2003 and MIPLIB2010. However, these algorithms fall short of consistently succeeding at the task due to an imbalance between exploration and exploitation. We propose BaMCTS, a Monte Carlo Tree Search framework for finding backdoors to MIPs. Extensive algorithmic engineering, hybridization with traditional MIP concepts, and close integration with the CPLEX solver have enabled our method to outperform baselines on MIPLIB2017 instances, finding backdoors more frequently and more efficiently. © 2021, CC BY.

关键词： Integer programming

Results of the Ontology Alignment Evaluation Initiative 2022 17

学校读者我要写书评

暂无评论

Results of the Ontology Alignment Evaluation Initiative 2022

17th International Workshop on Ontology Matching, OM 2022

作者： Pour, Mina Abd Nikooie Algergawy, Alsayed Buche, Patrice Castro, Leyla J. Chen, Jiaoyan Dong, Hang Fallatah, Omaima Faria, Daniel Fundulaki, Irini Hertling, Sven He, Yuan Horrocks, Ian Huschka, Martin Ibanescu, Liliana Jiménez-Ruiz, Ernesto Karam, Naouel Laadhar, Amir Lambrix, Patrick Li, Huanyu Li, Ying Michel, Franck Nasr, Engy Paulheim, Heiko Pesquita, Catia Saveta, Tzanina Shvaiko, Pavel Trojahn, Cassia Verhey, Chantelle Wu, Mingfang Yaman, Beyza Zamazal, Ondrej Zhou, Lu Linköping University Swedish E-Science Research Center Linköping Sweden Heinz Nixdorf Chair for Distributed Information Systems Friedrich Schiller University Jena Germany UMR IATE INRAE University of Montpellier France ZB MED Information Centre for Life Sciences Germany Department of Computer Science The University of Manchester United Kingdom Department of Computer Science University of Oxford United Kingdom Information School The University of Sheffield Sheffield United Kingdom University of Lisbon Portugal Institute of Computer Science-FORTH Heraklion Greece Data and Web Science Group University of Mannheim Germany Fraunhofer Institute for High-Speed Dynamics Ernst-Mach-Institut EMI Germany Université Paris-Saclay INRAE AgroParisTech UMR MIA Paris-Saclay France City University of London United Kingdom Fraunhofer FOKUS Institute for Applied Informatics University of Leipzig Germany University of Stuttgart Germany University of Gävle Sweden University Côte d'Azur CNRS Inria France Albert Ludwig University of Freiburg Germany LASIGE Faculdade de Ciências Universidade de Lisboa Portugal Trentino Digitale SpA Trento Italy Institut de Recherche en Informatique de Toulouse France World Data System International Technology Office United States Australian Research Data Commons Australia ADAPT Centre Trinity College Dublin Ireland Prague University of Economics and Business Czech Republic TigerGraph Inc. United States SIRIUS University of Oslo Norway

The Ontology Alignment Evaluation Initiative (OAEI) aims at comparing ontology matching systems on precisely defined test cases. These test cases can be based on ontologies of different levels of complexity and use different evaluation modalities. The OAEI 2022 campaign offered 14 tracks and was attended by 18 participants. This paper is an overall presentation of that campaign. © 2022 Copyright for this paper by its authors.

关键词： Ontology

Optimizing convolutional neural networks for Chronic Obstructive Pulmonary Disease detection in clinical computed tomography imaging

学校读者我要写书评

暂无评论

computers in Biology and Medicine 2025年 185卷 109533-109533页

作者： Dorosti, Tina Schultheiss, Manuel Hofmann, Felix Thalhammer, Johannes Kirchner, Luisa Urban, Theresa Pfeiffer, Franz Schaff, Florian Lasser, Tobias Pfeiffer, Daniela Chair of Biomedical Physics Department of Physics School of Natural Sciences Technical University of Munich Bavaria Garching85748 Germany Munich Institute of Biomedical Engineering Technical University of Munich Bavaria Garching85748 Germany Department of Diagnostic and Interventional Radiology School of Medicine and Health Klinikum rechts der Isar Technical University of Munich Bavaria Munich81675 Germany Institute for Advanced Study Technical University of Munich Bavaria Garching85748 Germany Computational Imaging and Inverse Problems Department of Computer Science School of Computation Information and Technology Technical University of Munich Bavaria Garching85748 Germany

We aim to optimize the binary detection of Chronic Obstructive Pulmonary Disease (COPD) based on emphysema presence in the lung with convolutional neural networks (CNN) by exploring manually adjusted versus automated window-setting optimization (WSO) on computed tomography (CT) images. 7194 contrast-enhanced CT images (3597 with COPD;3597 healthy controls) from 78 subjects were selected retrospectively (01.2018–12.2021) and preprocessed. For each image, intensity values were manually clipped to the emphysema window setting and a baseline ‘full-range’ window setting. Class-balanced train, validation, and test sets contained 3392, 1114, and 2688 images. The network backbone was optimized by comparing various CNN architectures. Furthermore, automated WSO was implemented by adding a customized layer to the model. The image-level area under the Receiver Operating Characteristics curve (AUC) [lower, upper limit 95% confidence] was utilized to compare model variations. Repeated inference (n = 7) on the test set showed that the DenseNet was the most efficient backbone and achieved a mean AUC of 0.80 [0.76, 0.85] without WSO. Comparably, with input images manually adjusted to the emphysema window, the DenseNet model predicted COPD with a mean AUC of 0.86 [0.82, 0.89]. By adding a customized WSO layer to the DenseNet, an optimal window in the proximity of the emphysema window setting was learned automatically, and a mean AUC of 0.82 [0.78, 0.86] was achieved. Detection of COPD with DenseNet models was improved by WSO of CT data to the emphysema window setting range. © 2024 The Authors

关键词： Lung cancer

GAIA: A Global, Multi-modal, Multi-scale Vision-Language Dataset for Remote Sensing Image Analysis

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Zavras, Angelos Michail, Dimitrios Zhu, Xiao Xiang Demir, Begüm Papoutsis, Ioannis Orion Lab School of Rural Surveying and Geoinformatics Engineering National Technical University of Athens Athens15772 Greece Institute of Astronomy Astrophysics Space Applications and Remote Sensing National Observatory of Athens Athens11810 Greece Department of Informatics and Telematics Harokopio University of Athens Athens17676 Greece Chair of Data Science in Earth Observation Technical University of Munich Munich80333 Germany Munich Center for Machine Learning Munich80333 Germany Faculty of Electrical Engineering and Computer Science Technische Universität Berlin Berlin10623 Germany BIFOLD - Berlin Institute for the Foundations of Learning and Data Berlin10623 Germany

The continuous operation of Earth-orbiting satellites generates vast and ever-growing archives of Remote Sensing (RS) images. Natural language presents an intuitive interface for accessing, querying, and interpreting the data from such archives. However, existing Vision-Language Models (VLMs) are predominantly trained on web-scraped, noisy image-text data, exhibiting limited exposure to the specialized domain of RS. This deficiency results in poor performance on RS-specific tasks, as commonly used datasets often lack detailed, scientifically accurate textual descriptions and instead emphasize solely on attributes like date and location. To bridge this critical gap, we introduce GAIA, a novel dataset designed for multi-scale, multi-sensor, and multi-modal RS image analysis. GAIA comprises of 205,150 meticulously curated RS image-text pairs, representing a diverse range of RS modalities associated to different spatial resolutions. Unlike existing vision-language datasets in RS, GAIA specifically focuses on capturing a diverse range of RS applications, providing unique information about environmental changes, natural disasters, and various other dynamic phenomena. The dataset provides a spatially and temporally balanced distribution, spanning across the globe, covering the last 25 years with a balanced temporal distribution of observations. GAIA’s construction involved a two-stage process: (1) targeted web-scraping of images and accompanying text from reputable RS-related sources, and (2) generation of five high-quality, scientifically grounded synthetic captions for each image using carefully crafted prompts that leverage the advanced vision-language capabilities of GPT-4o. We also release an automated processing framework developed for this purpose, enabling the broader research community to generate captions for RS images using the web-crawled image-text data. Our extensive experiments, including fine-tuning of CLIP and BLIP2 models, demonstrate that GAIA significan

关键词： Visual languages

Support Vector Machines with the Hard-Margin Loss: Optimal Training via Combinatorial Benders’ Cuts

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Santana, Ítalo Serrano, Breno Schiffer, Maximilian Vidal, Thibaut Department of Computer Science Pontifical Catholic University of Rio de Janeiro Brazil TUM School of Management Technical University of Munich Germany Munich Data Science Institute Technical University of Munich Germany CIRRELT Department of Mathematics and Industrial Engineering Polytechnique Montréal Canada SCALE-AI Chair in Data-Driven Supply Chains

The classical hinge-loss support vector machines (SVMs) model is sensitive to outlier observations due to the unboundedness of its loss function. To circumvent this issue, recent studies have focused on non-convex loss functions, such as the hard-margin loss, which associates a constant penalty to any misclassified or within-margin sample. Applying this loss function yields much-needed robustness for critical applications but it also leads to an NP-hard model that makes training difficult, since current exact optimization algorithms show limited scalability, whereas heuristics are not able to find high-quality solutions consistently. Against this background, we propose new integer programming strategies that significantly improve our ability to train the hard-margin SVM model to global optimality. We introduce an iterative sampling and decomposition approach, in which smaller subproblems are used to separate combinatorial Benders’ cuts. Those cuts, used within a branch-and-cut algorithm, permit to converge much more quickly towards a global optimum. Through extensive numerical analyses on classical benchmark data sets, our solution algorithm solves, for the first time, 117 new data sets to optimality and achieves a reduction of 50% in the average optimality gap for the hardest datasets of the benchmark. Copyright © 2022, The Authors. All rights reserved.

关键词： Support vector machines

Differentiation between maize and weed plants using multispectral UAV image data for subarea specific weed control 41

学校读者我要写书评

暂无评论

Differentiation between maize and weed plants using multispe...

41. Jahrestagung 2021 der Gesellschaft fur Informatik in der Land-, Forst- und Ernahrungswirtschaft (GIL) - Fokus: "Informations- und Kommunikationstechnologien in kritischen Zeiten" - 2021 41st Annual Meeting of the Society for computer science in Agriculture, Forestry and the Food Industry (GIL) - Focus: "Information and Communication Technologies in Critical Times"

作者： Wittstruck, Lucas Reuter, Tobias Nahrstedt, Konstantin Trautz, Dieter Kühling, Insa Jarmer, Thomas Universität Osnabrück Institute of Computer Science Remote Sensing Group Wachsbleiche 27 Osnabrück49090 Germany Hochschule Osnabrück University of Applied Sciences Faculty of Agricultural Sciences and Landscape Architecture Am Krümpel 31 Osnabrück49090 Germany Kiel University Agronomy and Crop Science Hermann-Rodewald-Str. 9 Kiel24118 Germany

Fast and efficient parallel breadth-first search with power-law graph transformation

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Jiang, Zite Liu, Tao Zhang, Shuai Guan, Zhen Yuan, Mengting You, Haihang School of Computer Science and Technology University of Chinese Academy of Sciences SKL Computer Architecture Institute of Computing Technology Chinese Academy of Sciences China SKL Computer Architecture Institute of Computing Technology Chinese Academy of Sciences China School of Computer Science Wuhan University China

In the big data era, graph computing is widely used to exploit the hidden value in real-world graphs in various scenarios such as social networks, knowledge graphs, web searching, and recommendation systems. However, the random memory accesses result in inefficient use of cache and the irregular degree distribution leads to substantial load imbalance. Breadth-First Search (BFS) is frequently utilized as a kernel for many important and complex graph algorithms. In this paper, we describe a preprocessing approach using Reverse Cuthill-Mckee (RCM) algorithm to improve data locality and demonstrate how to achieve an efficient load balancing for BFS. Computations on RCM-reordered graph data are also accelerated with SIMD executions. We evaluate the performance of the graph preprocessing approach on Kronecker graphs of the Graph500 benchmark and real-world graphs. Our BFS implementation on RCM-reordered graph data achieves 326.48 MTEPS/W (mega TEPS per watt) on an ARMv8 system, ranking 2nd on the Green Graph500 list in June 2020 (the 1st rank uses GPU acceleration). © 2020, CC BY.

关键词： Graphic methods