检索结果-内蒙古大学图书馆

4th International Conference on Computer Network, Electronic and Automation, ICCNEA 2021

作者： Wang, Xin Yan, Liting Zhang, Qizhi School Of Electrical Engineering Xi'an Shiyou University Xi'an China Shaanxi Provincial Key Lab Of Oil And Gas Well Measurement And Control Technology Xi'an Shiyou University Xi'an China

ISBN: (纸本)9781665444866

The gradient descent algorithm is a type of optimization algorithm that is widely used to solve machine learning algorithm model parameters. Through continuous iteration, it obtains the gradient of the objective function, gradually approaches the optimal solution of the objective function, and finally obtains the minimum loss function and related parameters. The gradient descent algorithm is frequently used in the solution process of logical regression, which is a common binary classification approach. This paper compares and analyzes the differences between batch gradient descent and its derivative algorithms - stochastic gradient descent algorithm and mini- batch gradient descent algorithm in terms of iteration number, loss function through experiments, and provides some suggestions on how to pick the best algorithm for the logistic regression binary task in machine learning. © 2021 IEEE.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Real-Time Routing Design for LEO Satellite Networks: An Enhanced Multi-Agent DRL Approach

Real-Time Routing Design for LEO Satellite Networks: An Enha...

引用

IEEE International Conference on communications in China Workshops (ICCC)

作者： Hailong Liao Xian Zhang Jiaen Zhou Xuehua Li Key Laboratory of Information and Communication Systems Ministry of Information Industry and the Key Laboratory of Modern Measurement & Control Technology School of Information and Communication Engineering Beijing Information Science and Technology University Beijing China State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China

ISBN: (数字)9798350377675

ISBN: (纸本)9798350377682

Low Earth Orbit (LEO) satellite networks have been heralded as a promising solution for provisioning space-based Internet services. However, the inherent dynamic topology, characterized by frequent variations in inter-satellite link states, causes pronounced fluctuations in latency and exacerbated packet loss rates. To address the challenge, in this paper, we formulate a routing design and optimization problem to maximize the average throughput with considering the end-to-end latency constraint and the real-time routing demand in LEO satellite networks. To solve this problem, we design an enhanced Multi-Agent Deep Reinforcement Learning (MADRL) algorithm, which encapsulates the intricate interplay among LEO satellite agents and their environs within a Real-Time Markov Decision Process (RTMDP) framework, and introduces the self-attention mech-anism to endow agents with a better grasp of environmental state dynamics. Simulation results demonstrate that our proposed routing design outperforms the benchmark schemes in terms of both end-to-end latency and network average throughput.

关键词： Satellites Network topology Simulation Heuristic algorithms Low earth orbit satellites Routing Throughput Real-time systems Topology Optimization

来源：评论

学校读者我要写书评

暂无评论

Improved Res2Net Based on Multi-scale Attention Mechanism for Breast Cancer Image Classification 12th

Improved Res2Net Based on Multi-scale Attention Mechanism fo...

引用

12th International Conference on Computer engineering and Networks, CENet 2022

作者： Lei, Suqin Deng, Xing Shao, Haijian Xia, Wang Department of Computer Jiangsu University of Science and Technology Jiangsu Zhenjiang212003 China College of Automation Key Laboratory of Ministry of Education of Complex Engineering System Measurement and Control Southeast University Jiangsu Nanjing210009 China School of Information Science and Technology Nantong University Nantong Jiangsu226019 China Department of Electrical and Computer Engineering National University of Singapore Singapore117583 Singapore

ISBN: (纸本)9789811969003

Breast cancer is one of the most common cancers among female diseases. Since the classification accuracy of pathological images is crucial to the diagnosis of breast cancer, in order to reduce the error of manual diagnosis and improve the automatic classification accuracy of pathological images, this paper proposes a high-accuracy breast cancer classification method based on multi-scale attention mechanism enhanced Res2Net. By constructing a multi-channel residual connection in the residual block, and introducing a convolutional attention mechanism to enhance the feature extraction ability of Res2Net, and using group convolution to replace part of the ordinary convolution. In addition, since the channel information interaction across groups cannot be realized in the group convolution, which reduces the feature extraction ability. In this paper, the channel shuffling operation is used to reorganize the feature maps after the group convolution to achieve the information interaction between different groups. The evaluation results on the BreaKHis dataset show that, the accuracy and precision of the method in this paper have reached 98.5% and 98.9% respectively, in binary classification of histopathological images compared with the benchmark method, and the advantages of recall rate and F1 value are also more obvious. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Deep Learning for 3D Point Cloud Enhancement: A Survey

arXiv

引用

arXiv 2024年

作者： Quan, Siwen Yu, Junhao Nie, Ziming Wang, Muze Feng, Sijia An, Pei Yang, Jiaqi School of Electronics and Control Engineering Chang’an University Xi’an710064 China School of Software Engineering Northwestern Polytechnical University Xi’an710064 China School of Computer Science Northwestern Polytechnical University Xi’an710064 China School of Electronic Information and Communications Huazhong University of Science and Technology China Wuhan National Laboratory for Optoelectronics Wuhan430074 China

Point cloud data now are popular data representations in a number of three-dimensional (3D) vision research realms. However, due to the limited performance of sensors and sensing noise, the raw data usually suffer from sparsity, noise, and incompleteness. This poses great challenges to down-stream point cloud processing tasks. In recent years, deep-learning-based point cloud enhancement methods, which aim to achieve dense, clean, and complete point clouds from low-quality raw point clouds using deep neural networks, are gaining tremendous research attention. This paper, for the first time to our knowledge, presents a comprehensive survey for deep-learning-based point cloud enhancement methods. It covers three main perspectives for point cloud enhancement, i.e., (1) denoising to achieve clean data;(2) completion to recover unseen data;(3) upsampling to obtain dense data. Our survey presents a new taxonomy for recent state-of-the-art methods and systematic experimental results on standard benchmarks. In addition, we share our insightful observations, thoughts, and inspiring future research directions for point cloud enhancement with deep learning. Copyright © 2024, The Authors. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Study on fast response of semiconductor laser driven by impulse current

Study on fast response of semiconductor laser driven by impu...

引用

2020 International Symposium on Electronic Information technology and Communication engineering, ISEITCE 2020

作者： Qing, Shi Gentai, Shi Tao, Zhang Zhongmin, Xing Department of Measurement Technology and Control Engineering School of Mechanical Engineering Sichuan University Chengdu610065 China

A simulation model based on single-mode rate equations of semiconductor laser and its typical parameters shows that, the photon stimulated emission rate at the early turn-on time of the semiconductor laser is increased with the increase of the injection current, and the corresponding rise time is reduced with the increase of injection current. However, for an output power is fixed semiconductor laser, increasing the injection current to reduce the rise time of the transmitted pulse leads to increase in temperature of semiconductor laser. We show that adding an impulse current in the front of the normal luminous pulse current can reduce the rise time of the transmitted pulse and ensure the stable output of the semiconductor laser. This method is verified by simulation and an experiment using PLTB450B semiconductor laser. The simulation and test results show that for an output power is fixed semiconductor laser, the rise time of the transmitted pulse can be reduced significantly by adding impulse current. © Published under licence by IOP Publishing Ltd.

关键词： Semiconductor lasers

来源：评论

学校读者我要写书评

暂无评论

The fabrication of "8"-shape microstructured optical fiber and its application in pressure/temperature sensing

引用

Journal of Lightwave technology 2025年

作者： Xue, Pingsheng Fu, Yongqing Wu, Qiang Ng, Wai Pang Binns, Richard Liu, Qiang Northeastern University College of Information Science and Engineering Shenyang110819 China Northumbria University Faculty of Engineering and Environment Newcastle upon TyneNE1 8ST United Kingdom Northeastern University at Qinhuangdao School of Control Engineering Qinhuangdao066004 China Hebei Key Laboratory of Micro-Nano Precision Optical Sensing and Measurement Technology Qinhuangdao066004 China

A microstructured optical fiber was designed. The internal structure of the fiber consists of two round channels, and its cross-section resembles the shape of "8". This fiber can be applied in the fields of pressure sensing and temperature sensing. The stress induced by pressure on the core increases the birefringence due to the photo-elastic effect. When an appropriate liquid material, such as glycerol, is filled into the two round channels of the fiber, temperature sensing can be achieved based on the coupling between the fiber core and the liquid rod. Both pressure and temperature sensing characteristics were investigated in detail in experiments. By constructing a Sagnac interferometer, the pressure sensitivity reached 109.78 nm/MPa. Based on the coupling between the core and the infiltrated glycerol in the round channels, the temperature sensitivity was as high as -47.15 nm/°C. Their cross-sensitivities were also analyzed. The pressure or temperature sensors based on this fiber are highly suitable for scenarios such as the oil and gas industries. © 1983-2012 IEEE.

关键词： Gas industry

来源：评论

学校读者我要写书评

暂无评论

Htmatch: An Efficient Hybrid Transformer Based Graph Neural Network for Local Feature Matching

SSRN

引用

SSRN 2022年

作者： Cai, Youcheng Li, Lin Wang, Dong Li, Xinjie Liu, Xiaoping School of Computer Science and Information Engineering Hefei University of Technology Hefei230009 China Engineering Research Center of Safety Critical Industrial Measurement and Control Technology Ministry of Education China

Local feature matching plays a vital role in various computer vision tasks. In this work, we present a novel network that combines feature matching and outlier rejection for finding reliable correspondences between image pairs. The proposed method is a hybrid Transformer based Graph Neural Network (GNN), termed HTMatch, which aims to achieve high accuracy and efficient feature matching. Specifically, we first propose a hybrid Transformer that integrates self- and cross-attention together to condition the feature descriptors between image pairs. By doing so, the intra/inter-graph attentional aggregation can be realized by a single Transformer layer, which achieves more efficient message passing. Then, we introduce a new spatial embedding module to enhance the spatial constraints across images. The spatial information from one image is embedded into another, which can significantly improve match performance. Finally, we adopt a seeded GNN architecture for establishing a sparse graph, which improves both efficiency and effectiveness. Experiments show that HTMatch reaches state-of-the-art results on several public benchmarks. © 2022, The Authors. All rights reserved.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Strain-regulated sensing properties of α-Fe_(2)O_(3) nano-cylinders with atomic carbon layers for ethanol detection

引用

Journal of Materials Science & technology 2021年第9期68卷 132-139页

作者： Huajing Xiong Jianan Fu Jinyao Li Rashad Ali Hong Wang Yifan Liu Hua Su Yuanxun Li Woon-Ming Lau Nasir Mahmood Chunhong Mu Xian Jian School of Materials and Energy University of Electronic Science and Technology of ChinaChengdu 611731China Institute of Chemistry National Institute of Measurement and Testing Technology of ChinaChengdu 610021China School of Electronic Science and Engineering University of Electronic Science and Technology of ChinaChengdu 611731China Jiangxi Guo Chuang Industrial Park Development Co. LtdGanzhou 341000China Institute of Fundamental and Frontier Science University of Electronic Science and Technology of ChinaChengdu 610054China School of Engineering RMIT University124 La Trobe Street3001 MelbourneVictoriaAustralia National Engineering Research Center of Electromagnetic Radiation Control Materials State Key Laboratory of Electronic Thin Films and Integrated DevicesUniversity of Electronic Science and Technology of ChinaChengdu 611731China

The sensitivity of ethanol sensor is of paramount importance in a variety of areas,including chemical production with ethanol,alcohol testing for driving safety,***,α-Fe_(2)O_(3) nano-cylinders with atomic carbon layers are synthesized,for the first time,through in-situ catalytic chemical vapor deposition combined with hydrothermal techniques for the detection of *** reported α-Fe_(2)O_(3)@C nano-cylinders with double surficial strain effects deliver an ethanol detection sensitivity of 8 times as compared with α-Fe_(2)O_(3) nano-cylinders,10 times higher as compared with its detection sensitivity to ammonia,para-xylene,methanol and *** sensor also exhibits over-14-day operation stability and the minimum detection limit of 10 *** our best knowledge,the performances surpass those of previously reported α-Fe_(2)O_(3).Such attractive performances are attributed to the enhanced charge transfer in α-Fe_(2)O_(3) owing to the double surficial strain effects of α-Fe_(2)O_(3)@C nano-cylinders and the efficient adsorption of ethanol with atomic carbon layers.

关键词： Gas sensor Nano-cylinder Hybrid microstructure Sensitivity Carbon coating

来源：评论

学校读者我要写书评

暂无评论

Optimizing Frequency Selection for Cole-Cole Model Parameter Extraction in Bioimpedance Spectroscopy Using CPE

Optimizing Frequency Selection for Cole-Cole Model Parameter...

引用

Impedance Spectroscopy (IWIS), International Workshop on

作者： Cherif Ouni Nour Ammar Ahmed Yahia Kallel John Horstmann Nabil Derbel Ahmed Fakhfakh Olfa Kanoun Measurement and Sensor Technology Faculty of Electrical Engineering and Information Technology Chemnitz University of Technology Chemnitz Germany Laboratory of Signals Systems Artificial Intelligence and Networks Digital Research Centre of Sfax National School of Electronics and Telecommunications of Sfax Sfax Tunisia Control and Energy Management Laboratory National Engineering School of Sfax University of Sfax Sfax Tunisia Chemnitz University of Technology Chemnitz Germany Electronic Devices of Micro and Nano Technique Faculty of Electrical Engineering and Information Technology Chemnitz University of Technology Chemnitz Germany

ISBN: (数字)9798331507626

ISBN: (纸本)9798331507633

Wearable bioimpedance spectroscopy (BIS) devices are becoming increasingly popular due to their portability and low power consumption. However, improving power efficiency remains a significant challenge. One proposed solution is to reduce the number of excitation frequencies in the stimulation signal. This study investigates how reducing the number of used frequencies affects bioimpedance spectroscopy performance through simulations using synthetic data derived from a Cole-Cole model incorporating a Constant Phase Element (CPE). Simulations were conducted at various noise levels to evaluate how noise impacts the relationship between frequency and impedance. The results indicate a correlation between the number of frequencies used in BIS and the accuracy of CPE parameter extraction. This is particularly evident for $\alpha$ parameters of 0.8 and higher. Conversely, lowering the $\alpha$ value increases errors, even when a higher number of frequencies is used.

关键词： Spectroscopy Accuracy Biological system modeling Noise Bioimpedance Frequency estimation Data models Impedance Parameter extraction Synthetic data

来源：评论

学校读者我要写书评

暂无评论

PCDH-based nonlinear H∞control of VSC-HVDC with wind power integrated

引用

Journal of control and Decision 2021年第1期8卷 41-49页

作者： Xingyu Lv Bangjun Lei Shumin Fei Foreign language Teaching Center Guizhou Institute of TechnologyGuiyangPeople’s Republic of China Key Laboratory of Power Dig Data of Guizhou Province School of Electrical and Information EngineeringGuizhou Institute of TechnologyGuiyang 550003People’s Republic of China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of EducationSchool of AutomationSoutheast UniversityNanjingPeople’s Republic of China

Based on the port-controlled dissipative Hamiltonian(PCDH)system,a novel nonlinear(NN)H∞control is designed for the wind farm side(WFS)voltage source converter(VSC)of VSC-based high-voltage direct current(VSC-HVDC)transmission system connected large wind ***,the uncertain nonlinear dynamical model of WFS VSC in the VSC-HVDC is established by taking external disturbances and parameter variations into ***,the dissipative Hamiltonian structure of the WFS VSC system is constructed by means of variable ***,on the basis of the obtained dissipative Hamiltonian structure,a NN H∞control is put ***,the simulation results show that the proposed NN H∞control gains the advantage over the interconnection and damping assignment passivity-based(IDA-PB)control.

关键词： VSC-HVDC large wind farm Hamiltonian function method nonlinear H∞control

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：