检索结果-内蒙古大学图书馆

7th international conference on Future Networks and distributed systems, ICFNDS 2023

作者： Nurnoby, M Faisal Shawarib, Khalid A. Abu Khan, Ayaz Ul Hassan King Fahd University Of Petroleum And Minerals Dhahran Saudi Arabia King Fahd University Of Petroleum And Minerals Khobar Saudi Arabia SDAIA-KFUPM Joint Research Center For Artificial Intelligence King Fahd University Of Petroleum And Minerals Dhahran Saudi Arabia

ISBN: (纸本)9798400709036

Artificial intelligence has shown great potential in a variety of applications, from natural language models to audio visual recognition, classification, and manipulation. AI Researchers have to work with massive amount of collected data for use in machine learning, raising some challenges in effectively managing and utilizing the collected data in the training phase to develop and iterate on more accurate, and more generalized models. In this paper we conducted a review on parallel and distributed machine learning methods and challenges. We also propose a distributed and scalable deep learning model architecture which can span across multiple processing nodes. We tested the model on the MIT Indoor dataset, to evaluate the performance and scalability of the model using multiple hardware nodes, and showed the scaling characteristics of the different model using different model sizes. We find that distributed training is 80% faster using 2 GPUs than 1 GPU. We also find that the model keeps the benefits of distributed training such as speed and accuracy regardless of its size or training batch size. © 2023 ACM.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Stability analysis and parameter configuration of parallel virtual synchronous generators in microgrid

引用

ENERGY REPORTS 2022年 8卷 866-873页

作者： Hu, Xuekai Fan, Hui Li, Tiecheng Meng, Zhengji Luo, Peng Zhou, Hao Hebei Elect Power Res Inst Shijiazhuang 050021 Hebei Peoples R China

The current small-signal stability analysis and parameter selection of converters that mimic synchronous generators usually ignore some dynamic characteristics and the changes of steady-state operating points for simplicity. It could only judge the trend of small-signal stability and the analysis result was not accurate enough. Firstly, this paper established a small-signal model that mimics synchronous generators by connecting two isolated converters in parallel, the eigenvalues of the system were calculated and some of their physical meanings were obtained. Following that, in order to demonstrate the relationship of eigenvalues and state variables, the participation factor was calculated. Considering that, there will be some variability of the operating points with the parameters variation. As a function of the system settings, numerous root locus plots were created. For individual eigenvalue close to the real axis, in order to reveal the contribution of different parameters to the oscillation patterns, a separate eigenvalue sensitivity analysis was performed. The rules of the influence are essential to guide the design of parameters and their optimization of isolated converters connected in parallel that mimic synchronous generators. (C) 2022 The Author(s). Published by Elsevier Ltd.

关键词： Small-signal model Parameter configuration Participation factor distributed generation Virtual synchronous generator

来源：评论

学校读者我要写书评

暂无评论

Connecting Domain Experts and Data: Enriching User-Centric Data Analysis with Neural Network-Aided Data Source Suggestion 25th

Connecting Domain Experts and Data: Enriching User-Centric D...

引用

25th international conference on Enterprise Information systems (ICEIS)

作者： Behringer, Michael Treder-Tschechlov, Dennis Voggesberger, Julius HirmePD, Pascal Mitschang, Bernhard Univ Stuttgart Inst Parallel & Distributed Syst Univ Str 38 D-70569 Stuttgart Germany

ISBN: (纸本)9783031647475;9783031647482

Nowadays, data analysis is widely used in numerous areas to identify new trends, opportunities, or risks and to improve decision-making. In many cases, however, data analysis is only possible by incorporating specific domain knowledge, which is why domain experts need to be involved. To this end, data mashups are a popular tool for modeling tailored analyses. Yet, with today's data volumes from heterogeneous source systems, it is very difficult to identify beneficial data sources, in particular for explorative data analysis. In this paper, we first define requirements aiming for user-centric analytics, followed by the introduction of SDRank, a deep-learning-based approach to identify beneficial data sources. In an extensive evaluation with three scenarios, we show that this approach offers high robustness concerning the training data used and can reliably identify beneficial data sources, even for previously unknown domains, i.e., transfer learning.

关键词： User-centric data analysis Data mashups

来源：评论

学校读者我要写书评

暂无评论

Joint international conference of the 13th IFToMM international Symposium on Science of Mechanisms and Machines, SYROM 2022 and 25th international conference on Robotics, ROBOTICS 2022

Joint International Conference of the 13th IFToMM Internatio...

引用

Joint international conference of the 13th IFToMM international Symposium on Science of Mechanisms and Machines, SYROM 2022 and 25th international conference on Robotics, ROBOTICS 2022

ISBN: (纸本)9783031256547

The proceedings contain 42 papers. The special focus in this conference is on Joint international conference of the international Symposium on Science of Mechanisms and Machines and international conference on Robotics. The topics include: Study of Vibrations of a Cantilever Beam with Uniformly distributed Mass;considerations on the Proper Selection of Sensors for Vibroacoustic Study of the Vehicles;experimental Identification of Singularities in parallel Manipulators;requirements and Problems for Space Berthing System;A Kinematic Analysis of the TORVEastro Astronaut Robot Arm;Evaluation by Numerical Simulation of Friction Forces in Spherical Joints of a 6-DOF parallel Topology Robot;comparative Analysis of Four Solutions for Command and Control Applied to Anthropomorphic Grippers for Robots;Model-Free Reaching of a 2-DOF Robotic Arm Using Neural Networks;inverse Dynamic Modeling of a parallel Robot for Lower Limb Rehabilitation;a Study of Feasibility of a Mechanism for Rib-Vertebra-Sternum Prosthesis;kinematic and Dynamic Analysis of an Exoskeleton Robotic System;design and Operation of a Cable-Driven Robot for Lower-Limb Rehabilitation;a State of Art Overwiew on Wrist Rehabilitation Exoskeletons;design of an Exoskeleton for Rehabilitation Ankle Joint;requirements and Characteristics of Finger Motion Assistance;development of a Force Feedback Control for Robotic Assisted Liver Cancer Treatment;Perspectives on Originally Designed Eco-Friendly Robotic Cell for PCB Dismantling;cloud-Based Digital Twin for Robot Health Monitoring and Integration in Cyber-Physical Production systems;a Novel Design of a Robotic System for Biological Fluid Aliquoting;multimodal Perceptual Cues for Context-Aware Human-Robot Interaction;a Review on Mechanisms Used for the Reconfigurable Wheelchairs.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Task-parallel Runtime for Heterogeneous Multi-node Vector systems 23rd

A Task-Parallel Runtime for Heterogeneous Multi-node Vector...

引用

23rd international conference on parallel and distributed Computing, Applications, and Technologies, PDCAT 2022

作者： Ide, Kazuki Takahashi, Keichi Shimomura, Yoichi Takizawa, Hiroyuki Graduate School of Information Sciences Tohoku University Miyagi Sendai980-8578 Japan Cyberscience Center Tohoku University Miyagi Sendai980-8578 Japan

ISBN: (纸本)9783031299261

In recent years, high-performance computing systems are equipped with not only host processors but also accelerators, and becoming more heterogeneous as well as becoming larger in scale. The task parallel execution model is promising to efficiently utilize such a large-scale system by minimizing synchronizations in comparison with traditional models. In this paper, we propose a task-parallel runtime system that individually considers the processors for task management and task execution;those two roles could be assigned to different processors. This paper focuses on NEC SX-Aurora TSUBASA as an example of heterogeneous multi-node systems, which are equipped with two kinds of general-purpose processors, to exploit the system heterogeneity for efficient task-parallel execution. Specifically, the proposed runtime system is used to select an appropriate processor for task management, depending on several execution conditions. The performance of the proposed runtime is discussed by running a Cholesky factorization implementation. The evaluation results show that the proposed runtime system can improve performance by more than 25% in comparison with a conventional a conventional implementation. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： parallel programming

来源：评论

学校读者我要写书评

暂无评论

Custom Accessors: Enabling Scalable Data Ingestion, (Re-)Organization, and Analysis on distributed systems

Custom Accessors: Enabling Scalable Data Ingestion, (Re-)Org...

引用

2024 IEEE international conference on Big Data, BigData 2024

作者： Castellana, Vito Giovanni Mutlu, Burcu O. Dio Lavore, Ian Di Firoz, Jesun Wolf, Katherine Minutoli, Marco Feo, John Pacific Northwest National Laboratory RichlandWA United States Politecnico di Milano Milano Italy

ISBN: (纸本)9798350362480

The emerging class of high velocity and high volume data analytic workflows comprise interwoven data ingestion, organization, and processing stages, with ingestion and organization steps often contributing comparable or even higher computational costs than actual processing steps. Since complex workflows consist of a variety of phases that view and use data differently, being able to construct efficient, scalable, distributed data structures (arrays, vectors, sets, maps, and multi-maps) is essential and requires custom methods to extend and shrink containers, analyze and position data, and, maintain globally-consistent meta-data. In this paper, we propose a novel data-structure access paradigm based on the concept of Accessors. At a high level, accessors are customizable callable objects that can modify the behavior of insert, read, update, and delete operations for distributed containers while preserving atomicity guarantees. Accessors provide a very clean and natural way to implement a variety of programming patterns, e.g., conditional insertion/deletion and cascading computations, which would be otherwise hard (or even impossible) to express in parallel and distributed settings without using locks. We demonstrate the practicality and usefulness of our approach with two representative use cases and study the performance of these applications on a distributed High-Performance Computing system. Our analysis highlights that our proposed abstraction allows for an effective overlapping and concurrent execution of different workflow steps (e.g., data ingestion and analysis), which in a conventional analytics pipeline would execute sequentially, contributing cumulatively to the overall latency. © 2024 IEEE.

关键词： C++ (programming language)

来源：评论

学校读者我要写书评

暂无评论

An Exploration of Global Optimization Strategies for Autotuning OpenMP-based Codes

An Exploration of Global Optimization Strategies for Autotun...

引用

1st international conference on Smart Energy systems and Artificial Intelligence (SESAI)

作者： Bolet, Gregory Georgakoudis, Giorgis Parasyris, Konstantinos Cameron, Kirk W. Beckingsale, David Gamblin, Todd Virginia Tech Blacksburg VA 24061 USA Lawrence Livermore Natl Lab Livermore CA 94550 USA

ISBN: (纸本)9798350364613;9798350364606

Automatic parameter tuning of parallel codes is ubiquitous in today's HPC environments where the performance portability of said codes is expected to keep pace with the perpetual release of new hardware. With changes in hardware, it is often the case that finding an optimal configuration of these codes is challenging and only further complicated by the high dimensionality or discontinuous topologies of the tuning spaces. Selecting a proper optimization strategy to automatically search these spaces is paramount to minimizing the energy and time spent on exploring sub-optimal configurations. Unfortunately, it is often the case that these optimizers have hyperparameters of their own, which are sensitive and can greatly affect the outcome of quickly converging to, or even finding an optimal code configuration. Much of the existing autotuning literature tends to use particular optimizers without describing their hyperparameter selection, leaving readers to figure out how to configure their optimizer for the best performance. In this work we compare and contrast the popular global optimization strategy of Bayesian Optimization (BO) to Iwo less popular strategies: Particle Swarm Optimization (PSO), and Covariance Matrix Adaptive Evolution Strategy (CMA-ES). We sweep the hyperparameters of these three optimizers in the context of tuning OpenMP hyperparameters of four classic OpenMP programs: BT, FT, HPCG, and Lulesh. Our study compares the long-term search behavior and average time-to-convergence between these three optimization strategies in tuning OpenMP codes. We contribute a detailed study of these strategies and provide deeper insights as to their sensitivities, noting the conditions where each performs well, and hinting at which optimizers require minimal tuning of their hyperparameters for desirable tuning results.

关键词： bayesian optimization particle swarm optimization covariance matrix adaptive evolution strategy openmp nas parallel benchmarks hpcg lulesh

来源：评论

学校读者我要写书评

暂无评论

Improved parallel Genetic Algorithm for Fixed Charge Transportation Problem

Improved Parallel Genetic Algorithm for Fixed Charge Transpo...

引用

international conference on Artificial Intelligence and Smart Environment, ICAISE 2022

作者： Lahjouji El Idrissi, Ahmed Ezzerrifi Amrani, Ismail El Allaoui, Ahmad Abdelmalek Essaâdi University Tetouan Morocco L-STI T-IDMS FST Errachidia Moulay Ismail University of Meknes Meknes Morocco

ISBN: (纸本)9783031262531

Genetic algorithm (GA) are a powerful method to different combinatorial and NP-Complete problems. This helps us find an optimal solution if the exact methods are not effective. Additionally, GA delivers the best solution in a reasonable time. But standard genetic algorithms require a rather high execution cost, especially for larger instances. However, the parallel genetic algorithm (PGA) is a mechanism that uses several genetic algorithms at the same time to solve the same task. In this work, we present a new approach to solve the FCTP problem using parallel genetic algorithms (PGA). Indeed, we propose a parallel algorithm-oriented multiprocessor architecture with distributed memory. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

Combining Lossy Compression with Multi-level Caching for Data Staging over Network

Combining Lossy Compression with Multi-level Caching for Dat...

引用

1st international conference on Smart Energy systems and Artificial Intelligence (SESAI)

作者： Aoyagi, Rei Takahashi, Keichi Shimomura, Yoichi Takizawa, Hiroyuki Tohoku Univ Grad Sch Informat Sci Sendai Miyagi Japan Tohoku Univ Cybersci Ctr Sendai Miyagi Japan

ISBN: (纸本)9798350364613;9798350364606

Researchers conduct post-processing on the simulation results by running an interactive data analysis tool on a High-Performance Computing (HPC) system installed at an HPC center and retrieving the post-processed results. Certain data analysis scenarios require to transfer the simulation results directly from the center. in such scenarios, a portion of the data would usually be streamed over the network to achieve interactivity. However, there still exist two challenges in maintaining interactivity: (1) limited network bandwidth and (2) long network latency. To tackle these challenges, we propose a system to enable interactive array analysis over the network. We employ error-bounded lossy compression to increase the effective network bandwidth. Furthermore, we employ multi-level caching to hide the network latency and combine prefetching to improve the cache hit ratio. The cache replacement and prefetching policies are designed considering the data access pattern of interactive analysis. We compared our proposed system with TileDB, one of the state-of-the-art array databases, by measuring the average latency for various access patterns. Compared to TileDB, the proposed system reduces the average latency by up to 91.6% by allowing 10% of error because the cache hit ratio was improved by more than 40% due to the proper cache replacement and prefetching policy and network transfer time was reduced more than 75% by using lossy compression.

关键词： HPC distributed Computing Data Staging Lossy Compression Interactive Analysis Bandwidth Software Caching

来源：评论

学校读者我要写书评

暂无评论

FreezePipe: An Efficient Dynamic Pipeline parallel Approach Based on Freezing Mechanism for distributed DNN Training 26

FreezePipe: An Efficient Dynamic Pipeline Parallel Approach ...

引用

26th international conference on Computer Supported Cooperative Work in Design, CSCWD 2023

作者： Weng, Caishan Shu, Zhiyang Xu, Zhengjia Zhang, Jinghui Luo, Junzhou Dong, Fang Wang, Peng Wang, Zhengang Southeast University School of Computer Science and Engineering Nanjing China Boston College Nanjing China Co. Ltd China Jiangsu HHigh-Tech Software Technology Co.Ltd China

ISBN: (纸本)9798350331684

Deep Neural Network (DNN) training on a large scale is extremely time-consuming and computationally intensive, which is accelerated by distributed training. In recent years, pipeline parallelism has been developed, which enables partitioning the model across several devices, e.g. GPU, and training efficiency is improved by dividing data batches into micro-batches, with each of them processed by a different stage of the model. Currently, parallel training assumes pipeline placement and partitioning are static, with parameters updating each iteration, without accounting for freezing. This results in computational resources not being fully utilized. In this paper, we propose FreezePipe, a novel method for optimizing deep learning training that combines the freezing mechanism with pipeline parallel training. In FreezePipe, a lightweight method for determining the freezing strategy based on gradient changes is employed. Considering that resources need to be released based on the frozen layer, a lightweight model partitioning algorithm was designed to determine the optimal strategy for pipeline partitioning. Experimental results show that FreezePipe can reduce the training time by 64.5% compared to Torchgpipe on CIFAR-10 dataset without compromising any model performance. © 2023 IEEE.

关键词： Pipelines

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：