检索结果-内蒙古大学图书馆

20th European Conference on Computer Systems, EuroSys 2025, co-located 30th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, ASPLOS 2025

作者： Pan, Lichen Liu, Juncheng Fu, Yongquan Yuan, Jinhui Zhang, Rongkai Li, Pengze Xiao, Zhen School of Computer Science Peking University China OneFlow Research China National Key Laboratory of Parallel and Distributed Computing College of Computer Science and Technology National University of Defense Technology China

ISBN: (纸本)9798400711961

distributed deep neural network training necessitates efficient GPU collective communications, which are inherently susceptible to deadlocks. GPU collective deadlocks arise easily in distributed deep learning applications when multiple collectives circularly wait for each other. GPU collective deadlocks pose a significant challenge to the correct functioning and efficiency of distributed deep learning, and no general effective solutions are currently available. Only in specific scenarios, ad-hoc methods, making an application invoke collectives in a consistent order across GPUs, can be used to prevent circular collective dependency and deadlocks. This paper presents DFCCL, a novel GPU collective communication library that provides a comprehensive approach for GPU collective deadlock prevention while maintaining high performance. DFCCL achieves preemption for GPU collectives at the bottom library level, effectively preventing deadlocks even if applications cause circular collective dependency. DFCCL ensures high performance with its execution and scheduling methods for collectives. Experiments show that DFCCL effectively prevents GPU collective deadlocks in various situations. Moreover, extensive evaluations demonstrate that DFCCL delivers performance comparable to or superior to NCCL, the state-of-the-art collective communication library highly optimized for NVIDIA GPUs. © 2025 Copyright held by the owner/author(s).

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Partial Order-centered Hyperbolic Representation Learning for Few-shot Relation Extraction 31

Partial Order-centered Hyperbolic Representation Learning fo...

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Hu, Biao Huang, Zhen Hu, Minghao Yang, Pinglv Qiao, Peng Dou, Yong Wang, Zhilin National Key Laboratory of Parallel and Distributed Computing National University of Defense Technology China Center of Information Research Academy of Military Science China College of Meteorology and Oceanology National University of Defense Technology China

ISBN: (纸本)9798891761964

Prototype network-based methods have made substantial progress in few-shot relation extraction (FSRE) by enhancing relation prototypes with relation descriptions. However, the distribution of relations and instances in distinct representation spaces isolates the constraints of relations on instances, making relation prototypes biased. In this paper, we propose an end-to-end partial order-centered hyperbolic representation learning (PO-HRL) framework, which imposes the constraints of relations on instances by modeling partial order in hyperbolic space, so as to effectively learn the distribution of instance representations. Specifically, we develop the hyperbolic supervised contrastive learning based on Lorentzian cosine similarity to align representations of relations and instances, and model the partial order by constraining instances to reside within the Lorentzian entailment cone of their respective relation. Experiments on three benchmark datasets show that PO-HRL outperforms the strong baselines, especially in 1-shot settings lacking relation descriptions. © 2025 Association for Computational Linguistics.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

PEbfs: Implement High-Performance Breadth-First Search on PEZY-SC3s 24th

PEbfs: Implement High-Performance Breadth-First Search on P...

引用

24th International Conference on Algorithms and Architectures for parallel Processing, ICA3PP 2024

作者： Guo, Weihao Wang, Qinglin Liu, Xiaodong Peng, Muchun Yang, Shun Liang, Yaling Shi, Yongzhen Cao, Ligang Liu, Jie Laboratory of Digitizing Software for Frontier Equipment National University of Defense Technology Changsha410073 China National Key Laboratory of Parallel and Distributed Computing National University of Defense Technology Changsha410073 China Engineering Research Center for National Fundamental Software National University of Defense Technology Changsha410073 China

ISBN: (纸本)9789819615506

The breadth-first search (BFS) algorithm is a fundamental algorithm in graph theory, and it’s parallelization can significantly improve performance. Therefore, there have been numerous efforts to leverage the powerful parallel computing capabilities of hardware like GPGPU to implement high-performance BFS algorithms. However, the energy efficiency is relatively low due to the high power consumption of the platforms on which the algorithm is adapted to. To deal with these challenges, this paper introduces PEbfs that is a high-performance BFS algorithm based on the PEZY-SC3s efficient processor. We integrated three search algorithms, two algorithm optimization strategies, and a directional optimization scheme into PEbfs. Through multiple evaluations of the performance of PEbfs on the public SNAP dataset, the results demonstrate that the average energy efficiency ratio of PEbfs is higher than that of Enterprise and Tigr, the two most advanced implementations on Nvidia’s GPGPU: It achieves 3.08× the average energy efficiency ratio of Enterprise and 4.53× that of Tigr. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Optimization algorithms

来源：评论

学校读者我要写书评

暂无评论

Multi-class contour preserving classification

Multi-class contour preserving classification

引用

13th International Conference on Intelligent Data Engineering and Automated Learning, IDEAL 2012

作者： Fuangkhon, Piyabute Tanprasert, Thitipong Distributed and Parallel Computing Research Laboratory Faculty of Science and Technology Assumption University Bangkok Thailand

ISBN: (纸本)9783642326387

The original contour preserving classification technique was proposed to improve the robustness and weight fault tolerance of a neural network applied with a two-class linearly separable problem. It was recently found to be improving the level of accuracy of two-class classification. This paper presents an augmentation of the original technique to improve the level of accuracy of multi-class classification by better preservation of the shape or distribution model of a multi-class problem. The test results on six real world multi-class datasets from UCI machine learning repository present that the proposed technique supports multi-class data and can improve the level of accuracy of multi-class classification more effectively. © 2012 Springer-Verlag.

关键词： Neural networks

来源：评论

学校读者我要写书评

暂无评论

An End-to-End Classification Network Model Based on Hybrid Supervision for Industrial Surface Defect Detection

引用

International Journal of Pattern Recognition and Artificial Intelligence 2025年第5期39卷

作者： Qin, Runbing Chen, Ningjiang Gan, Shukun Guangxi University School of Computer Electronics and Information Nanning530004 China Education Department of Guangxi Zhuang Autonomous Region Key Laboratory of Parallel Distributed and Intelligent Computing Guangxi University Nanning530004 China Guangxi Intelligent Digital Services Research Center of Engineering Technology Nanning530004 China

Industrial part surface defect detection aims to precisely locate defects in images, which is crucial for quality control in manufacturing. The traditional method needs to be designed in advance, but it has shortcomings in terms of generalization ability. Deep neural network-based defect detection faces issues such as low resolution, data scarcity, labeling costs, and high computation. Therefore, an improved solution is necessary to enhance industrial product quality control efficiency and accuracy. A new method is proposed that utilizes end-to-end training of a two-stage neural network based on segmentation with an extended training process. The gradient flow from the classification to the segmentation network is adjusted to prevent unstable features from interfering with learning. To address the problem of image oversampling and undersampling during training, a frequency sampling scheme for negative samples is introduced. Additionally, positive pixels in the region-based segmentation mask are weighted using the distance transform algorithm so that regions with a high probability of defects can be detected without detailed annotation. Experiments are conducted across three distinct defect datasets by applying hybrid supervision encompassing diverse conditions. In the optimal case, the AP rate of the proposed model on the DAGM and KolektorSDD datasets reaches 100%, and the AP rate on the Severstal Steel dataset reaches 98.99%. The experiments’ results show that the detection accuracy has improved greatly, showing how well the proposed method works in the industry. © 2025 World Scientific Publishing Company.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

PDRL-CM: An efficient cooperative caching management method for vehicular networks based on deep reinforcement learning

引用

Ad Hoc Networks 2025年 176卷

作者： Pingjie Ou Ningjiang Chen Long Yang School of Computer and Electronic Information Guangxi University Nanning 530004 Guangxi China Key Laboratory of Parallel Distributed and Intelligent Computing Education Department of Guangxi Zhuang Autonomous Region Nanning 530004 Guangxi China The Guangxi Intelligent Digital Services Research Center of Engineering Technology Nanning 530004 Guangxi China

In vehicular networks, onboard devices face the challenge of limited storage, and computational resources constrain their processing and storage capabilities. This limitation is particularly significant for applications that require complex computations and real-time responses. Additionally, limited storage capacity reduces the range of cacheable data, which can impact the immediate availability of data and the continuity of services. Therefore, improving cache utilization and meeting vehicles’ real-time data demands pose significant challenges. Deep reinforcement learning can optimize the issues arising from agents’ continuously changing state and action spaces due to increasing request demands. However, training the network may encounter instability and convergence difficulties in dynamic and complex environments or situations with sparse rewards. In response to these issues, this paper proposes a Priority-based Deep Reinforcement Learning Collaborative Cache Management method (PDRL-CM). PDRL-CM first designs a lightweight cache admission strategy that leverages data’s inherent and combined attributes. It then makes cache admission decisions Using Monte Carlo sampling and a max-value search strategy combined with a feedforward neural network. Secondly, the method considers minimizing system latency and reducing vehicle energy consumption as joint optimization problems. An improved deep reinforcement learning algorithm solves this problem and makes cache-sharding decisions. A prioritized experience replay mechanism is incorporated to adjust the network prediction model quickly and accelerate the convergence process. Experimental results indicate that, compared to existing DRL-based caching methods, PDRL-CM offers faster data processing efficiency and higher cache hit rates under varying vehicle density, storage capacity, and content volume conditions.

关键词：

来源：评论

学校读者我要写书评

暂无评论

KC-GenRe: A Knowledge-constrained Generative Re-ranking Method Based on Large Language Models for Knowledge Graph Completion 30

KC-GenRe: A Knowledge-constrained Generative Re-ranking Meth...

引用

Joint 30th International Conference on Computational Linguistics and 14th International Conference on Language Resources and Evaluation, LREC-COLING 2024

作者： Wang, Yilin Hu, Minghao Huang, Zhen Li, Dongsheng Yang, Dong Lu, Xicheng Defense Innovation Institute Academy of Military Sciences China Information Research Center of Military Science China National Key Laboratory of Parallel and Distributed Computing China

ISBN: (纸本)9782493814104

The goal of knowledge graph completion (KGC) is to predict missing facts among entities. Previous methods for KGC re-ranking are mostly built on non-generative language models to obtain the probability of each candidate. Recently, generative large language models (LLMs) have shown outstanding performance on several tasks such as information extraction and dialog systems. Leveraging them for KGC re-ranking is beneficial for leveraging the extensive pre-trained knowledge and powerful generative capabilities. However, it may encounter new problems when accomplishing the task, namely mismatch, misordering and omission. To this end, we introduce KC-GenRe, a knowledge-constrained generative re-ranking method based on LLMs for KGC. To overcome the mismatch issue, we formulate the KGC re-ranking task as a candidate identifier sorting generation problem implemented by generative LLMs. To tackle the misordering issue, we develop a knowledge-guided interactive training method that enhances the identification and ranking of candidates. To address the omission issue, we design a knowledge-augmented constrained inference method that enables contextual prompting and controlled generation, so as to obtain valid rankings. Experimental results show that KG-GenRe achieves state-of-the-art performance on four datasets, with gains of up to 6.7% and 7.7% in the MRR and Hits@1 metric compared to previous methods, and 9.0% and 11.1% compared to that without re-ranking. Extensive analysis demonstrates the effectiveness of components in KG-GenRe. © 2024 ELRA Language Resource Association: CC BY-NC 4.0.

关键词： Knowledge graph

来源：评论

学校读者我要写书评

暂无评论

An incremental learning algorithm for supervised neural network with contour preserving classification

An incremental learning algorithm for supervised neural netw...

引用

International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology, ECTI-CON

作者： Piyabute Fuangkhon Thitipong Tanprasert Parallel and Distributed Computing Research Laboratory Faculty of Science and Technology Assumption University Bangkok Thailand

This paper presents an alternative algorithm for integrating the existing knowledge of a supervised learning neural network with the new training data. The algorithm allows the existing knowledge to age out in slow rate as a neural network is gradually retrained with consecutive sets of new samples, resembling the change of application locality under a consistent environment. The algorithm also utilizes the contour preserving classification algorithm to increase the accuracy of classification. The experiment is performed on 2-dimension partition problem and the result convincingly confirms the effectiveness of the algorithm.

关键词： Neural networks Support vector machines Support vector machine classification Supervised learning Partitioning algorithms Neurons Classification algorithms Speech recognition Data mining distributed computing

来源：评论

学校读者我要写书评

暂无评论

An Outpost Vector placement evaluation of an incremental learning algorithm for Support Vector Machine

An Outpost Vector placement evaluation of an incremental lea...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Piyabute Fuangkhon Thitipong Tanprasert Distributed and Parallel Computing Research Laboratory Faculty of Science and Technology Assumption University Bangkok Thailand

Outpost Vector model synthesizes new vectors at the boundary of two classes of data in order to increase the level of accuracy of classification. This paper presents a performance evaluation of four different placements of outpost vectors in an incremental learning algorithm for Support Vector Machine (SVM) on a non-complex problem. The algorithm generates outpost vectors from selected new samples, selected prior samples, both samples, or generates no outpost vector at all. After that, they are included in the final training set, as well as new samples and prior samples, based on the specified parameters. The experiments are conducted with a 2-dimension partition problem. The distribution of training and test samples is created in a limited location of a 2-dimension donut ring. There are two classes of data which are represented as 0 and 1. The context of the problem is assumed to shift 45 degrees in counterclockwise direction. Every consecutive partition is set to have different class of data. The experimental results show that the placement of outpost vectors generated from only selected new samples yields the highest level of accuracy of classification for both new data and old data. As a result, using samples from different part of the algorithm to generate outpost vectors affects the level of accuracy of classification.

关键词： Training Support vector machines Accuracy Shape Data models Context Vectors

来源：评论

学校读者我要写书评

暂无评论

A 55 TFLOPS simulation of amyloid-forming peptides from yeast prion Sup35 with the special-purpose computer system MDGRAPE-3

A 55 TFLOPS simulation of amyloid-forming peptides from yeas...

引用

2006 ACM/IEEE Conference on Supercomputing, SC'06

作者： Narumi, Tetsu Ohno, Yousuke Okimoto, Noriaki Koishi, Takahiro Suenaga, Atsushi Futatsugi, Noriyuki Yanai, Ryoko Himeno, Ryutaro Fujikawa, Shigenori Taiji, Makoto Ikei, Mitsuru Intel K.K. High Performance Molecular Simulation Team Genomic Sciences Center Innovative Nanopatterning Laboratory Topochemical Design Laboratory Frontier Research System Advanced Center for Computing and Communication Software and Solutions Group Parallel and Distributed Solutions Group

ISBN: (纸本)0769527000

We have achieved a sustained performance of 55 TFLOPS for molecular dynamics simulations of the amyloid fibril formation of peptides from the yeast Sup35 in an aqueous solution. For performing the calculations, we used the MDGRAPE-3 system - -a special-purpose computer system for molecular dynamics simulations. Its nominal peak performance was 415 TFLOPS for Coulomb force calculations;this is the highest-ever performance reported for classical molecular dynamics simulations. Amyloid fibril formation is known to be related to the occurrence of severe diseases such as Alzheimer's, Parkinson's, and Creutzfeldt-Jakob diseases. The Sup35 protein is a "yeast prion protein," which forms mini-crystals due to aggregation;it forms an effective platform for studying the formation process of amyloid fibrils. In these simulations, we first elucidate that the amyloid-forming peptides GNNQQNY aggregate at a higher frequency than non-amyloid-forming peptides SQNGNQQRG;further, the GNNQQNY peptides tend to form parallel two-stranded -sheets that would grow into a cross- amyloid nucleus. The results are consistent with those obtained experimentally. Furthermore, we could observe an early elongation of the amyloid nucleus. This result is expected to contribute toward a deeper understanding of the amyloid growth mechanism. © 2006 IEEE.

关键词： Peptides

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：