检索结果-内蒙古大学图书馆

2nd International Conference on Artificial Intelligence and computer science, AICS 2020

作者： Zheng, Ling Ye, Xin Chen, Fei Department of Computer Science and Technology North China Electric Power University Faculty of Control and Computer Engineering room E1112 No. 2 Beinong Road Huilongguan Town Changping District Beijing China

In order to improve the reliability and accuracy of the main steam temperature trend prediction, a main steam temperature prediction model based on improved LSTM is proposed. Firstly, uses the grey correlation analysis method to select the important influencing factors. Then, a linear structure is introduced into the LSTM structure to construct a main steam temperature prediction model. Finally, based on the historical operating data of the thermal power unit, a simulation experiment is performed to compare the prediction error of the output of the RNN model, the LSTM model, and the improved LSTM model. The results show that the method has higher prediction accuracy for the main steam temperature. At the same time, compared with other traditional methods, this method has better fitting effect, which can be well applied in practical engineering. © Published under licence by IOP Publishing Ltd.

关键词： Correlation methods

来源：评论

学校读者我要写书评

暂无评论

Intelligent gradient amplification for deep neural networks

arXiv

引用

arXiv 2023年

作者： Basodi, Sunitha Pusuluri, Krishna Xiao, Xueli Pan, Yi Georgia State University Georgia Institute of Technology Emory University 55 Park Place AtlantaGA30303 United States Department of Computer Science Georgia State University 25 Park Place NE AtlantaGA30303 United States Faculty of Computer Science and Control Engineering Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Shenzhen University Town Room D513 1068 Xueyuan Avenue Shenzhen518055 China

Deep learning models offer superior performance compared to other machine learning techniques for a variety of tasks and domains, but pose their own challenges. In particular, deep learning models require larger training times as the depth of a model increases, and suffer from vanishing gradients. Several solutions address these problems independently, but there have been minimal efforts to identify an integrated solution that improves the performance of a model by addressing vanishing gradients, as well as accelerates the training process to achieve higher performance at larger learning rates. In this work, we intelligently determine which layers of a deep learning model to apply gradient amplification to, using a formulated approach that analyzes gradient fluctuations of layers during training. Detailed experiments are performed for simpler and deeper neural networks using two different intelligent measures and two different thresholds that determine the amplification layers, and a training strategy where gradients are amplified only during certain epochs. Results show that our amplification offers better performance compared to the original models, and achieves accuracy improvement of ∼2.5% on CIFAR-10 and ∼4.5% on CIFAR-100 datasets, even when the models are trained with higher learning rates. © 2023, CC BY-NC-ND.

关键词： Backpropagation

来源：评论

学校读者我要写书评

暂无评论

Refinement of Massey Inequality : (To Nicolae Ţǎpuş on his 70th birthday)

Refinement of Massey Inequality : (To Nicolae Ţǎpuş on hi...

引用

2019 IEEE International Symposium on Information Theory, ISIT 2019

作者： Popescu, Pantelimon George Omar Choudary, Marios Computer Science and Engineering Department Faculty of Automatic Control and Computers University Politehnica of Bucharest Splaiul Independenţei 313 Bucharest060042 Romania

ISBN: (纸本)9781538692912

关键词： Information theory

来源：评论

学校读者我要写书评

暂无评论

Krasovskii and Shifted Passivity Based Output Consensus

arXiv

引用

arXiv 2022年

作者： Kawano, Yu Cucuzzella, Michele Feng, Shuai Scherpen, Jacquelien M.A. Graduate School of Advanced Science and Engineering Hiroshima University Higashi-Hiroshima739-8527 Japan Department of Electrical Computer and Biomedical Engineering University of Pavia PV Pavia27100 Italy School of Automation Nanjing University of Science and Technology Nanjing210094 China Jan C. Willems Center for Systems and Control ENTEG Faculty of Science and Engineering University of Groningen Groningen9747 AG Netherlands

Motivated by current sharing in power networks, we consider a class of output consensus (also called agreement) problems for nonlinear systems, where the consensus value is determined by external disturbances, e.g., power demand. This output consensus problem is solved by a simple distributed output feedback controller if a system is either Krasovskii or shifted passive, which is the only essential requirement. The effectiveness of the proposed controller is shown in simulation on an islanded DC power network. Copyright © 2022, The Authors. All rights reserved.

关键词： Nonlinear systems

来源：评论

学校读者我要写书评

暂无评论

Securing isosceles triangular formations under heterogeneous sensing and mixed constraints

arXiv

引用

arXiv 2021年

作者： Chan, Nelson P.K. Jayawardhana, Bayu de Marina, Hector Garcia Engineering and Technology Institute Groningen University of Groningen Nijenborgh 4 Groningen9747AG Netherlands Department of Computer Architecture and Automatic Control Faculty of Physics Universidad Complutense de Madrid Madrid28040 Spain

This paper focuses on securing a triangular shape (up to translation) for a team of three mobile robots that uses heterogeneous sensing mechanism. Based on the available local information, each robot employs the popular gradient-based control law to attain the assigned individual task(s). In the current work, robots are assigned either distance and signed area task(s) or bearing task(s). We provide a sufficient condition on the gain ratio RAd between the signed area and the distance control term such that the desired formation shape, an isosceles triangle, is reached from all feasible starting positions. Numerical simulations are provided to support the theoretical *** Codes 93D99 © 2021, CC BY-NC-SA.

关键词： Robots

来源：评论

学校读者我要写书评

暂无评论

A Horizontal Federated Learning Approach to IoT Malware Traffic Detection: An Empirical Evaluation with N-BaIoT Dataset

A Horizontal Federated Learning Approach to IoT Malware Traf...

引用

International Conference on Advanced Communication Technology (ICACT)

作者： Phuc Hao Do Tran Duc Le Vladimir Vishnevsky Aleksandr Berezkin Ruslan Kirichek The Bonch-Bruevich Saint-Petersburg State University of Telecommunications Saint-Petersburg Russia Danang Architecture University Da Nang Viet Nam Department of Communication Networks and Data Transmission at Bonch-Bruevich Saint-Petersburg State University of Telecommunications Russia University of Science and Technology - The University of Danang Da Nang Viet Nam Information Technology Faculty The University of Danang - University of Science and Technology Danang Vietnam V. A. Trapeznikov Institute of Control Sciences of Russian Academy of Sciences Moscow Russia Institute of Control Sciences of Russian Academy of Sciences Russia Department of Programming Engineering and Computer Science Bonch Bruevich Saint Petersburg State University of Telecommunications

The increasing prevalence of botnet attacks in IoT networks has led to the development of deep learning techniques for their detection. However, conventional centralized deep learning models pose challenges in simultaneously ensuring user data privacy and detecting botnet attacks. To address this issue, this study evaluates the efficacy of Federated Learning (FL) in detecting IoT malware traffic while preserving user privacy. The study employs N-BaIoT, a dataset of real-world IoT network traffic infected by malware, and compares the effectiveness of FL models using Convolutional Neural Network, Long Short-Term Memory, and Gated Recurrent Unit models with a centralized approach. The results indicate that FL can achieve high performance in detecting abnormal traffic in IoT networks, with the CNN model yielding the best results among the three models evaluated. The study recommends the use of FL for IoT malware traffic detection due to its ability to preserve data privacy.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Study of Structure Evolution of Ultra-Rapidly Annealed Fe75.3ni10b14cu0.7 Alloys

SSRN

引用

SSRN 2023年

作者： Kolano – Burian, Aleksandra Kowalczyk, Maciej Grabias, Agnieszka Blyskun, Piotr Gazinska, Malgorzata Zackiewicz, Przemyslaw Wójcik, Anna Chulist, R. Maziarz, Wojciech Pilśniak, Adam Bernacki, Krzysztof Faculty of Materials Science and Engineering Warsaw University of Technology Woloska 141 Warszawa05-507 Poland Lukasiewicz Research Network Institute of Non-Ferrous Metals Sowińskiego 5 Gliwice44-100 Poland Lukasiewicz Research Network Institute of Microelectronics and Photonics al. Lotników 32/46 Warszawa02-668 Poland The Aleksander Krupkowski Institute of Metallurgy and Materials Science Polish Academy of Sciences 25 Reymonta St. Kraków30-059 Poland Faculty of Chemistry Wroclaw University of Science and Technology Norwida 4/6 Wroclaw50-373 Poland Faculty of Automatic Control Electronics and Computer Science Silesian University of Technology Akademicka 16 Gliwice44-100 Poland

The article concerns modern soft magnetic materials made by rapid solidification and subsequent ultra-rapid annealing. Due to the lower amount of non-magnetic elements, higher magnetisation is possible. Unfortunately, the lack of those elements causes difficulty during isothermal nanocrystallisation operation, being crucial from the final product properties point of *** of isothermal heat treatment, ultra-rapid annealing is implemented. It reflects in materials with good magnetic softness and higher possible magnetisation *** amorphous alloy with Fe75.3Ni10B14Cu0.7 chemical composition was heat treated between cooper blocks at 773K in the range of 1-60s. To study structure evolution, we brought XRD, DSC, Flash DSC, Mössbauer spectroscopy, TEM, and HRTEM techniques to *** on the DSC results, we estimated the minimum heating speed to obtain optimal structure at 6000K/min. The Mössbauer results allowed us to study annealing time-dependent local atomic ordering in the amorphous state. A comparison of X-Ray and synchrotron diffraction results drove us to the conclusion that X-Ray measurements are sufficient to determine the material structure. TEM shows the formation of α-Fe crystallites even after an annealing time of 0.5s. The increase in annealing time leads to an increase in the crystalline phase volume fraction and growth of α-Fe crystallite size. © 2023, The Authors. All rights reserved.

关键词： Differential scanning calorimetry

来源：评论

学校读者我要写书评

暂无评论

Defective Sic Nanotube Based Single-Atom Catalysts for Electrocatalytic Nitrogen Fixation with Curvature Effect

SSRN

引用

SSRN 2023年

作者： Wang, Zhiwei Gao, Shuaishuai Liu, Xiaojing Chen, Xin Zhang, Xintao Sa, Rongjian Li, Qiaohong Sun, Chenghua Ma, Zuju School of Environmental and Materials Engineering Yantai University Yantai264005 China School of Computer and Control Engineering Yantai University Yantai264005 China College of Materials and Chemical Engineering Minjiang University Fuzhou350108 China State Key Laboratory of Structural Chemistry Fujian Institute of Research the Structure of Matter Chinese Academy of Sciences Fuzhou350002 China Department of Chemistry and Biotechnology Faculty of Science Engineering & Technology Swinburne University of Technology HawthornVIC3122 Australia

The production of fertilizers, explosives, and various other chemicals relies heavily on ammonia (NH3). Electrochemical reduction of nitrogen to ammonia, also known as the nitrogen reduction reaction (NRR), is a potential alternative to the Haber-Bosch process that can be carried out at ambient conditions. However, many NRR catalysts suffer from low selectivity and require a large overpotential to drive the reaction, limiting their practical application. In this study, we used density functional theory (DFT) calculations to investigate the mechanistic aspects of the NRR and the effect of curvature variation on the catalytic activity of silicon carbide nanotubes (SiCNTs) decorated with 3d-5dtransition metal single-atom catalysts. Our results suggest that the Os@SiCNT catalyst exhibits promising activity as an eNRR catalyst with an overpotential of 0.47 V and high selectivity over hydrogen evolution reaction (HER) competition. The electronic "acceptance-donation" interaction between N2 and Os promotes the adsorption and activation of N2 molecules on the catalyst surface. Additionally, our findings demonstrate that as the nanotube diameter decreases, the energy barrier for the NRR reaction decreases as well. However, this decrease in diameter also leads to a weaker capacity for *N2 adsorption, which may ultimately reduce the selectivity of the NRR reaction. These findings offer valuable insights that can aid in the rational design of nanotube-based electrocatalysts for eNRR. © 2023, The Authors. All rights reserved.

关键词： Density functional theory

来源：评论

学校读者我要写书评

暂无评论

Transition from time-variant to static networks: Timescale separation in N-intertwined mean-field approximation of susceptible-infectious-susceptible epidemics

引用

Physical Review E 2024年第3期109卷 034308-034308页

作者： Robin Persoons Mattia Sensi Bastian Prasse Piet Van Mieghem Faculty of Electrical Engineering Mathematics and Computer Science Delft University of Technology P.O. Box 5031 2600 GA Delft The Netherlands MathNeuro Team Inria at Université Côte d'Azur 2004 Rte des Lucioles 06410 Biot France Department of Mathematical Sciences “G. L. Lagrange” Politecnico di Torino Corso Duca degli Abruzzi 24 10129 Torino Italy European Centre for Disease Prevention and Control (ECDC) Gustav III's Boulevard 40 169 73 Solna Sweden

We extend the N-intertwined mean-field approximation (NIMFA) for the susceptible-infectious-susceptible (SIS) epidemiological process to time-varying networks. Processes on time-varying networks are often analyzed under the assumption that the process and network evolution happen on different timescales. This approximation is called timescale separation. We investigate timescale separation between disease spreading and topology updates of the network. We introduce the transition times T̲(r) and T¯(r) as the boundaries between the intermediate regime and the annealed (fast changing network) and quenched (static network) regimes, respectively, for a fixed accuracy tolerance r. By analyzing the convergence of static NIMFA processes, we analytically derive upper and lower bounds for T¯(r). Our results provide insights and bounds on the time of convergence to the steady state of the static NIMFA SIS process. We show that, under our assumptions, the upper-transition time T¯(r) is almost entirely determined by the basic reproduction number R0 of the network. The value of the upper-transition time T¯(r) around the epidemic threshold is large, which agrees with the current understanding that some real-world epidemics cannot be approximated with the aforementioned timescale separation.

关键词： Epidemic Spreading Erdos-Renyi graph Evolving networks Social networks

来源：评论

学校读者我要写书评

暂无评论

Swin Transformer with Spatial and Local Context Augmentation for Enhanced Semantic Segmentation of Remote Sensing Images

IEEE Open Journal of Signal Processing

引用

IEEE Open Journal of Signal Processing 2025年 6卷 608-620页

作者： Ding, Rong-Xing Xu, Yi-Han Yu, Gang Zhou, Wen Zhou, Ding Nanjing Forestry University College of Information Science and Technology Artificial Intelligence Nanjing210037 China Qinghai University School of Computer Technology and Applications Xining810016 China The University of Sheffield Department of Electronic and Electrical Engineering SheffieldS10 2TN United Kingdom University of Surrey 5G/6G Innovation Centre SurreyGU2 7XH United Kingdom Guangzhou Maritime University College of Low Altitude Equipment and Intelligent Control Guangzhou510725 China University of Malaya Faculty of Engineering Kuala Lumpur50603 Malaysia

Semantic segmentation of remote sensing images is extensively used in crop cover and type analysis, and environmental monitoring. In the semantic segmentation of remote sensing images, owning to the specificity of remote sensing images, not only the local context is required, but also the global context information makes an important role in it. Inspired by the powerful global modelling capability of Swin Transformer, we propose the LSENet network, which follows the encoder-decoder architecture of the UNet network. In encoding phase, we propose spatial enhancement module (SEM), which helps Swin Transformer further enhance feature extraction by encoding spatial information. In decoding stage, we propose local enhancement module (LEM), which is embedded in the Swin Transformer to improve the Swin Transformer to assist the network to obtain more local semantic information so as to classify pixels more accurately, especially in the edge region, the adding of LEM enables to obtain smoother edges. The experimental results on the Vaihingen and Potsdam datasets demonstrate the effectiveness of our proposed method. Specifically, the mIoU metric is 78.58% on the Potsdam dataset, 72.59% on the Vaihingen dataset and 64.49% on the OpenEarthMap dataset. © 2020 IEEE.

关键词： Semantic Segmentation

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：