检索结果-内蒙古大学图书馆

A comprehensive survey on graph neural network accelerators

Frontiers of computer Science 2025年第2期19卷 11-29页

作者： Jingyu LIU Shi CHEN Li SHEN School of Computer National University of Defense TechnologyChangsha 410073China Key Laboratory of Advanced Microprocessor Chips and Systems Changsha 410073China

Deep learning has gained superior accuracy on Euclidean structure data in neural *** a result,nonEuclidean structure data,such as graph data,has more sophisticated structural information,which can be applied in neural networks as well to address more complex and practical ***,actual graph data obeys a power-law distribution,so the adjacent matrix of a graph is random and *** processing accelerator(GPA)is designed to handle the problems ***,graph computing only processes 1-dimensional *** graph neural networks(GNNs),graph data is ***,GNNs include the execution processes of both traditional graph processing and neural network,which have irregular memory access and regular computation,*** obtain more information in graph data and require better model generalization ability,the layers of GNN are deeper,so the overhead of memory access and computation is *** present,GNN accelerators are designed to deal with this *** this paper,we conduct a systematic survey regarding the design and implementation of GNN ***,we review the challenges faced by GNN accelerators,and existing related works in detail to process ***,we evaluate previous works and propose future directions in this booming field.

关键词： graph neural network accelerators graph convolutional networks design space exploration deep learning domain-specific architecture

来源：评论

学校读者我要写书评

暂无评论

Multi-Scale Time Series Segmentation Network Based on Eddy Current Testing for Detecting Surface Metal Defects

引用

IEEE/CAA Journal of Automatica Sinica 2025年第3期12卷 528-538页

作者： Xiaorui Li Xiaojuan Ban Haoran Qiao Zhaolin Yuan Hong-Ning Dai Chao Yao Yu Guo Mohammad S.Obaidat George Q.Huang the School of Intelligence Science and Technology University of Science and Technology Beijing the Beijing Advanced Innovation Center for Materials Genome Engineering the Key Laboratory of Intelligent Bionic Unmanned Systems and the Institute of Materials Intelligent Technology Liaoning Academy of Materials IEEE the Department of Computer Science Hong Kong Baptist University the School of Computer and Communication Engineering Key Laboratory of Advanced Materials and Devices for Post-Moore Chips Ministry of Education University of Science and Technology Beijing the Beijing Advanced Innovation Center for Materials Genome Engineering University of Science and Technology Beijing the School of Computer and Communication Engineering University of Science and Technology Beijing the King Abdullah Ⅱ School of Information Technology The University of Jordan the Department of Computational Intelligence the School of Computing SRM University the School of Engineering The Amity University The Hong Kong Polytechnic University

In high-risk industrial environments like nuclear power plants, precise defect identification and localization are essential for maintaining production stability and safety. However, the complexity of such a harsh environment leads to significant variations in the shape and size of the defects. To address this challenge, we propose the multivariate time series segmentation network(MSSN), which adopts a multiscale convolutional network with multi-stage and depth-separable convolutions for efficient feature extraction through variable-length templates. To tackle the classification difficulty caused by structural signal variance, MSSN employs logarithmic normalization to adjust instance distributions. Furthermore, it integrates classification with smoothing loss functions to accurately identify defect segments amid similar structural and defect signal subsequences. Our algorithm evaluated on both the Mackey-Glass dataset and industrial dataset achieves over 95% localization and demonstrates the capture capability on the synthetic dataset. In a nuclear plant's heat transfer tube dataset, it captures 90% of defect instances with75% middle localization F1 score.

关键词： Eddy current testing nondestructive testing semantic segmentation time series analysis

来源：评论

学校读者我要写书评

暂无评论

SSC: An SRAM-Based Silence Computing Design for On-chip Memory 24th

SSC: An SRAM-Based Silence Computing Design for On-chip Mem...

引用

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

作者： Chen, Ziming Deng, Quan Hu, Yiyue He, Xiaowei Huang, Libo Wang, Yongwen College of Computer National University of Defense Technology Changsha China Key Laboratory of Advanced Microprocessor Chips and Systems Changsha China

ISBN: (纸本)9789819615445

The rapid development of emerging intelligent applications leads to a surge in computational demands and memory capacity requirements. Compute-in-memory (CIM) is a promising paradigm to alleviate the data movement bottleneck of emerging intelligent applications. SRAM-based CIM technology is employed to enhance the performance of general-purpose processors (e.g., CPU) by reusing on-chip memory (e.g. Cache) for computational tasks. In recent SRAM-based CIM works, peripheral compute circuits, e.g., adder trees, are introduced to improve system performance, as SRAM cells are hard to support arithmetic operations efficiently. However, the excessive peripheral circuit with a large area overhead degrades the memory density, which breaks the balance of computation and memory. To improve the computing capability and memory capability simultaneously of digital-based CIM (DCIM), we propose SSC, an SRAM-based silence computing CIM design, which leverages logic-in-memory operations and peripheral circuits to achieve parallel computing within the SRAM array. We propose 8+T (8T, 9T, and 11T) SRAM bitcells to support majority-of-three (MAJ3), COPY, and NOT logic, which is used to achieve a 3:2 compressor. Furthermore, a silence computing design, facilitated by a local connection between SRAM cells, is proposed to improve system parallelism, which supports parallel logic-in-memory operations and memory access. Our experiment results show that SSC achieves 3.7× memory density improvement compared to prior DCIMs normalized to 12nm. Additionally, the SWaP (TOPS/W×Kb/mm2) figure-of-merit emphasizes the importance of memory density and energy efficiency, showing that this work achieves a higher SWaP of 119.89, which is 1.84× more than prior DCIMs normalized to 12 nm. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： computer circuits

来源：评论

学校读者我要写书评

暂无评论

Nanoimprint lithography-assisted block copolymer self-assembly for hyperfine fabrication of magnetic patterns based on L10-FePt nanoparticles

引用

Science China Chemistry 2025年第5期68卷 2027-2034页

作者： Jie Zhang Yi Chen Guijun Li Zelin Sun Zhengong Meng Wai-Yeung Wong Department of Applied Biology and Chemical Technology and Research Institute for Smart Energy The Hong Kong Polytechnic University The Hong Kong Polytechnic University Shenzhen Research Institute Division of Integrative Systems and Design and Department of Electronic and Computer EngineeringThe Hong Kong University of Science and Technology School of Flexible Electronics (Future Technology) and Institute of Advanced Materials (IAM) Nanjing Tech University

L10-FePt-type bit-patterned media has provided a promising alternative for ultrahigh-density magnetic recording systems in the current digital era, but rapid fabrication of magnetic patterns with hyperfine bit islands is still challenging, especially with the target for miniaturization and scalable production simultaneously. Herein, Fe,Pt-containing block copolymers were utilized as single-source precursors for solution-processable patterning and subsequent generation of the demanding magnetic FePt dots by in situ pyrolysis. High-throughput nanoimprint lithography was initially employed to fabricate the predefined bit cells precisely,and then the intrinsic self-assembly of phase-separated block copolymers further drove the formation of accurate bit *** from the synergistic effect of top-down lithographic approach and bottom-up self-assembly, the customizable patterns could be achieved for large-scale mass production in targeted areas, but high-density isolated dots could also be accurately aligned along the patterned features after subsequent self-assembly. This reliable strategy would provide a good avenue to precisely construct ultrahigh-density magnetic data storage devices.

关键词： organometallic precursor magnetic nanoparticles block copolymer self-assembly nanoimprint lithography

来源：评论

学校读者我要写书评

暂无评论

Integrating GIS-Remote Sensing: A Comprehensive Approach to Predict Oceanographic Health and Coastal Dynamics

引用

Remote Sensing in Earth systems Sciences 2025年第1期8卷 200-212页

作者： Krishnamoorthy, R. Tanaka, Kazuaki Begum, M. Amina Centre for Advanced Wireless Integrated Technology Chennai Institute of Technology Chennai India Department of Computer Science and Systems Engineering Kyushu Institute of Technology Fukuoka Japan Department of ECE Sri Venkateswaraa College of Technology Tamilnadu Chennai India

The monitoring of oceanographic and coastal dynamics is essential for understanding the effects of climate change, predicting natural disasters, and managing coastal resources. Remote sensing technology, particularly through satellite imagery, has revolutionized the ability to collect vast amounts of environmental data over large spatial and temporal scales. However, the complexity of these datasets, which often include intricate spatial patterns and long-term temporal trends, requires advanced machine learning models capable of handling both dimensions effectively. In this study, we present a Hybrid GIS-Transformer Network for analyzing oceanographic and coastal dynamics using remote sensing data. The proposed model integrates the spatial feature extraction capabilities of the GIS network with the temporal analysis strengths of the transformer model. This approach enables a comprehensive understanding of both spatial and temporal patterns in high-resolution satellite imagery, allowing for more accurate detection and prediction of environmental changes. Our methodology was tested across multiple tasks, including coastal erosion detection, sea surface temperature (SST) prediction, wave height prediction, and ocean currents detection. The Hybrid GIS-Transformer model consistently outperformed existing models such as CNN, ResNet50, MobileNet, EfficientNet, and LSTM. For coastal erosion detection, the hybrid model achieved an accuracy of 96.12%, while for SST prediction, the accuracy was 95.43%. In wave height prediction, the model achieved an accuracy of 93.02%, and for ocean currents detection, an accuracy of 91.88% was recorded. These results demonstrate the model’s superiority in capturing both fine spatial details and long-term temporal dynamics compared with individual models. © The Author(s), under exclusive licence to Springer Nature Switzerland AG 2025.

关键词： Coastal dynamics Deep learning Environmental monitoring Hybrid GIS-Transformer Machine learning Oceanographic analysis Remote sensing Sea surface temperature prediction Spatial feature extraction Temporal analysis Wave height prediction

来源：评论

学校读者我要写书评

暂无评论

EFS-XGBoost: A robust framework for precision classification of COVID-19 cases

引用

Intelligent Data Analysis 2025年第1期29卷 171-185页

作者： Abidi, Mustufa Haider Khare, Neelu D, Preethi Alkhalefah, Hisham Umer, Usama Advanced Manufacturing Institute King Saud University Riyadh Saudi Arabia School of Computer Science Engineering and Information Systems Vellore Institute of Technology Tamil Nadu Vellore India Department of Computer Science and Engineering Faculty of Engineering and Technology SRM Institute of Science and Technology Ramapuram Chennai India

The emergence of the novel COVID-19 virus has had a profound impact on global healthcare systems and economies, underscoring the imperative need for the development of precise and expeditious diagnostic tools. Machine learning techniques have emerged as a promising avenue for augmenting the capabilities of medical professionals in disease diagnosis and classification. In this research, the EFS-XGBoost classifier model, a robust approach for the classification of patients afflicted with COVID-19 is proposed. The key innovation in the proposed model lies in the Ensemble-based Feature Selection (EFS) strategy, which enables the judicious selection of relevant features from the expansive COVID-19 dataset. Subsequently, the power of the eXtreme Gradient Boosting (XGBoost) classifier to make precise distinctions among COVID-19-infected patients is *** EFS methodology amalgamates five distinctive feature selection techniques, encompassing correlation-based, chi-squared, information gain, symmetric uncertainty-based, and gain ratio approaches. To evaluate the effectiveness of the model, comprehensive experiments were conducted using a COVID-19 dataset procured from Kaggle, and the implementation was executed using Python programming. The performance of the proposed EFS-XGBoost model was gauged by employing well-established metrics that measure classification accuracy, including accuracy, precision, recall, and the F1-Score. Furthermore, an in-depth comparative analysis was conducted by considering the performance of the XGBoost classifier under various scenarios: employing all features within the dataset without any feature selection technique, and utilizing each feature selection technique in isolation. The meticulous evaluation reveals that the proposed EFS-XGBoost model excels in performance, achieving an astounding accuracy rate of 99.8%, surpassing the efficacy of other prevailing feature selection techniques. This research not only advances the field of COVI

关键词： COVID-19

来源：评论

学校读者我要写书评

暂无评论

CoDPoC IP: A Configurable Data Protection Circuit to Support Multiple Key Agreement Scheme 26th

CoDPoC IP: A Configurable Data Protection Circuit to Suppor...

引用

26th International Conference on Information and Communications Security, ICICS 2024

作者： Peng, Yijing Wang, Junhui Wang, Zhenyu Guo, Zhenbin Wang, Yaohua Deng, Ding Li, Shaoqing Guo, Yang Key Laboratory of Advanced Microprocessor Chips and Systems National University of Defense Technology Changsha China School of Computer National University of Defense Technology Changsha China College of Electronic Science and Technology National University of Defense Technology Changsha China College of Semiconductors Hunan University Changsha China

ISBN: (纸本)9789819788002

Physical Unclonable Function(PUF)-based security protocols are effective in mitigating various security threats, including eavesdropping, interception, and cloning. Nevertheless, most existing protocols are implemented in software, with the primary drawback being their inability to integrate with actual PUF and their inefficiency. Although some strategies incorporate PUF into their design, these systems do not allow for configuration changes and are deficient in flexibility. In this paper, we propose the CoDPoC IP, a configurable data protection circuit that supports PUF-based various key agreement authentication schemes and allows for flexible scheme design adjustments through configurable registers. The CoDPoC IP offers two advantages: 1) hardware-level implementation incorporates PUF, leading to higher resource utilization, efficiency, and security;and 2) the IP supports flexible configuration of schemes. Experimental results indicate that the hardware execution time is 0.03504 milliseconds. The proposed scheme is successfully implemented and verified with low-overhead hardware. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Data privacy

来源：评论

学校读者我要写书评

暂无评论

An Empirical Comparison of Machine and Deep Learning Algorithms for Predicting Maternal Health Risk 5th

An Empirical Comparison of Machine and Deep Learning Algorit...

引用

5th EAI International Conference on Cognitive Computing and Cyber Physical systems, IC4S 2024

作者： N, Srinivasa Gupta Valarmathi, B. Kavitha, A. Gopini, Sai Bhuvan Seela, Hari Kumar Department of Manufacturing Engineering School of Mechanical Engineering Vellore Institute of Technology Tamil Nadu Vellore India Department of Software and Systems Engineering School of Computer Science Engineering and Information Systems Vellore Institute of Technology Tamil Nadu Vellore India Department of Mathematics School of Advanced of Sciences Vellore Institute of Technology Tamil Nadu Vellore India

ISBN: (纸本)9783031770777

Maternal health is among the greatest challenges in the world, especially in rural areas as there lack medical practitioners, they do not have easily accessible publics clinics and transport is difficult. Therefore, high rates of maternal as well as infant morbidity and mortalities are recorded. This research utilizes Artificial Intelligence (AI) with machine learning algorithms to forecast and address maternal health hazards right at their onset stage. The current research utilizes the concept of AI along with many Machine Learning (ML) methods like the Ensemble Learning Model (ELM), Random Forest (RF), K-Nearest Neighbour (KNN), Decision-Tree (DT), XG-Boost (XGB), Cat Boost (CB), and Gradient Boosting (GB), along with Synthetic Minority Over-sampling Technique (SMOTE) algorithm used for dealing with the problem class imbalance within the data set. SMOTE algorithm is utilized for the dataset balancing process. The handling system involves refining data preprocessing with the help of feature engineering and robust data cleaning which makes sure that anomalies do not erode the reliability of the predictive model. The existing methods [1] used RF (90%), DT (87%), XGB (85%), CB (86%), and GB (81%) algorithms and were compared with the accuracies of the proposed models like Logistic Regression (LR), Ensemble Learning Bagging (ELB), Ensemble Learning Stacking (ELS), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU). The existing methods used only imbalance dataset. The accuracies of the proposed models with using SMOTE algorithm (balanced dataset) are LR (61.33%), KNN (81%), ELB (92.33%), ELS (90.66%) CNN (40.67%), RNN (59.67%), LSTM (54%), GRU (56%) respectively. Among these methods, ELB achieved 92.33% of accuracy with using SMOTE algorithm using imbalanced dataset. Whereas the accuracies of the proposed models without using SMOTE algorithm (imbalanced dataset) are LR (66.09%), KNN (68.47%)

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

MIN: Multi-stage Interactive Network for Multimodal Recommendation 25th

MIN: Multi-stage Interactive Network for Multimodal Recomme...

引用

25th International Conference on Web Information systems Engineering, WISE 2024

作者： Mo, Minghao Lu, Weihai Xie, Qixiao Lv, Xiang Xiao, Zikai Yang, Hong Zhang, Yanchun Cyberspace Institute of Advanced Technology Guangzhou University Guangzhou China School of Computer Science and Technology Zhejiang Normal University Zhejiang China Peking University Beijing China Department of Industrial and Systems Engineering The Hong Kong Polytechnic University Hung Hom China

ISBN: (纸本)9789819605699

With the surge in multimedia applications and the explosion of multimedia information, multimodal recommendation has garnered increasing attention. The core idea of multimodal recommendation is to effectively utilize the multimodal information of items to enhance the performance of recommendation systems. Previous work has primarily focused on integrating item ID embeddings and multimodal representations. However, there are two main drawbacks: first, a lack of interaction between multimodal features, which neglects the inherent semantic relationships within these features. Second, the distribution gap between ID embeddings and multimodal representations adversely affects recommendation performance. In this paper, we propose a novel Multi-stage Interactive Network for multimodal recommendation, named MIN. Specifically, we first discarded the ID embeddings, retaining only the multimodal features to represent the items. Then, we designed a three-stage interaction network to model the multimodal representations and the latent structural semantics between users and items. Additionally, to obtain more general and robust representations, we applied graph enhancement techniques in the final stage to enhance the representations of users and items. Finally, contrastive learning techniques were employed to address the inconsistency in user and item representations. A series of comprehensive experiments on three publicly available datasets have validated the efficacy of our method. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

A Novel Thermal Analysis Method for Tubular PM Linear Motors Based on Transfer Learning

引用

IEEE Transactions on Transportation Electrification 2025年第3期11卷 7379-7388页

作者： Wu, Tao Dai, Peipei Xue, Gang Guo, Youguang Lei, Gang Zhu, Jianguo Wang, Yifei China University of Geosciences School of Automation Wuhan430074 China Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Wuhan430074 China University of Technology Sydney School of Electrical and Data Engineering NSW2007 Australia University of Sydney School of Electrical and Computer Engineering SydneyNSW2006 Australia

Thermal modeling and analysis are critical for permanent magnet linear synchronous motors (PMLSMs), particularly in multi-physical analysis and motor design optimization. This paper proposes a new method for thermal steady state modeling and analysis based on transfer learning, which combines the thermal mechanism model with the data-driven model to achieve high-precision temperature analysis with small number of finite element analysis (FEA) samples. First, an equivalent thermal circuit (ETC) model is established for PMLSMs based on the structural parameters and heat transfer principles. Second, a hybrid model based on transfer learning is developed: the model is pre-trained by a large sample set generated from the ETC in the source field (mechanism analytical model), and then the trained model is transferred to the target field (FEA data-driven model). The parameters of the neural network in the target field are optimized by a few FEA sample sets with higher precision to improve the prediction accuracy of the target model. Experimental results demonstrate that the proposed method significantly boosts the performance of the deep neural network (DNN) model, particularly when training sets are small. The root mean square error (RMSE) of the DNN model decreases by 0.901 in the 5:5 (training/testing) dataset and 1.726 in the 2:8 dataset;and the corresponding R2 and maximal absolute error of motor performances are significantly decreased. The smaller the proportion of the training set, the better performance of the hybrid modeling will achieve. Furthermore, a temperature rise experiment conducted on a prototype motor validates the efficacy and precision of the proposed hybrid modeling methodology. © 2015 IEEE.

关键词： Permanent magnets

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：