检索结果-内蒙古大学图书馆

M22: A Communication-Efficient Algorithm for Federated Learning Inspired by Rate-Distortion

IEEE Transactions on Communications 2024年第2期72卷 845-860页

作者： Liu, Yangyi Rini, Stefano Salehkalaibar, Sadaf Chen, Jun McMaster University Department of Electrical and Computer Engineering HamiltonONL8S 4L8 Canada Department of Electrical and Computer Engineering Hsinchu30010 Taiwan University of Toronto Department of Electrical and Computer Engineering TorontoONM5A 4T4 Canada

In federated learning (FL), the communication constraint between the remote clients and the Parameter Server (PS) is a crucial bottleneck. For this reason, model updates must be compressed so as to minimize the loss in accuracy resulting from the communication constraint. This paper proposes 'M-magnitude weighted L2 distortion + 2 degrees of freedom' (M22) algorithm, a rate-distortion inspired approach to gradient compression for federated training of deep neural networks (DNNs). In particular, we propose a family of distortion measures between the original gradient and the reconstruction we referred to as ' M-magnitude weighted L{2} ' distortion, and we assume that gradient updates follow an i.i.d. distribution-generalized normal or Weibull, which have two degrees of freedom. In both the distortion measure and the gradient distribution, there is one free parameter for each that can be fitted as a function of the iteration number. Given a choice of gradient distribution and distortion measure, we design the quantizer to minimize the expected distortion in gradient reconstruction. To measure the gradient compression performance under a communication constraint, we define the per-bit accuracy as the optimal improvement in accuracy that one bit of communication brings to the centralized model over the training period. Using this performance measure, we systematically benchmark the choice of gradient distribution and distortion measure. We provide substantial insights on the role of these choices and argue that significant performance improvements can be attained using such a rate-distortion inspired compressor. © 1972-2012 IEEE.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

MultiClass Classification of Chest Diseases using CXR Images with DenseNet201+CNN and Grad CAM Visualization 2

MultiClass Classification of Chest Diseases using CXR Images...

引用

2nd International Conference on Power, electrical, Electronics and Industrial Applications, PEEIACON 2024

作者： Ahad, Tanzim Kibria, Hafsa Binte Mehemud, Md. Yusuf Rajshahi University of Engineering & Technology Department of Electrical & Computer Engineering Rajshahi6204 Bangladesh

ISBN: (纸本)9798331517984

Accurately identifying lung diseases is really important for treating them well, especially now with COVID-19 making respiratory problems worse worldwide. In response to the global impact, this paper has presented a deep learning framework for the multiclass classification of chest diseases, covering COVID-19, pulmonary fibrosis, tuberculosis, lung opacity, and normal cases. A modified DenseNet201 combined with a custom CNN has been introduced in this paper for a more effective classification of lung diseases. The proposed DenseNet201+CNN network surpasses key transfer learning models (ResNet50, VGG16, VGG19, DenseNet201) achieving a testing accuracy of 95.43%. This study employs DenseNet201 due to its superior feature extraction capabilities and efficiency in handling vanishing gradients. Contributions of this study include a 5-class classification system with a sizable combined dataset, addressing class imbalance through augmentation and rigorous model evaluation comparing with four other transfer learning models. The incorporation of Explainable AI (Artificial Intelligence), specifically Gradient Weighted Class Activation Mapping (Grad-CAM), enhances data visualization, ensuring transparency in the model's decision-making process. This study strives to advance automated disease diagnosis, potentially transforming early detection approaches for critical lung conditions. © 2024 IEEE.

关键词： COVID-19

来源：评论

学校读者我要写书评

暂无评论

A novel on–off linear quadratic regulator control approach for satellite rendezvous

引用

Aerospace Systems 2023年第4期6卷 613-620页

作者： Rouzegar, Hossein Ghanbarisabagh, Mohammad Department of Electrical and Computer Engineering Semnan University Semnan Iran Department of Electrical Engineering Faculty of Electrical Engineering and Computer Sciences Islamic Azad University North Tehran Branch Tehran Iran

In this paper, a novel on–off linear quadratic regulator (LQR) control for satellite rendezvous as an example of linear systems with on–off inputs has been proposed for the first time. It simultaneously benefits from unique potentials of LQR control method and the extensive applications of systems with on–off inputs in various areas. The on–off LQR control approach has been applied on the system of orbital rendezvous and docking of satellites equipped with thrusters which are appropriate samples of systems with on–off inputs. Because of the energy consumption significance in many practical applications, the proposed approach is designed to consume less energy as well. Simulation results show the energy consumption of the presented method has been reduced about 36% compared to the continuous LQR approach. © 2023, Shanghai Jiao Tong University.

关键词： Satellites

来源：评论

学校读者我要写书评

暂无评论

A Survey on Transformer-Based Extractive Summarization Methods 19

A Survey on Transformer-Based Extractive Summarization Metho...

引用

19th Iranian Conference on Intelligent Systems, ICIS 2024

作者： Afsharizadeh, Mahsa Mousavirad, Seyed Jalaleddin Sirjan University of Technology Department of Electrical and Computer Engineering Sirjan Iran Mid Sweden University Department of Computer and Electrical Engineering Sundsvall Sweden

ISBN: (纸本)9798331507565

Searching for a topic on the internet exposes the user to a large amount of textual data. Reviewing all these documents is not feasible in terms of time. Text summarization is one of the active areas in the field of natural language processing. This area is divided into two main categories: extractive and abstractive. Extractive summarization identifies informative and salient sentences from the text and outputs them as the result. Abstractive summarization paraphrases the text and generates a summary with new sentences. Various methods have been proposed for both types so far, including statistical methods, machine learning-based methods, deep learning, and more recently, transformer models. The transformer model is a neural network-based architecture that has achieved promising results in various tasks in natural language processing, including extractive summarization. This paper provides a comprehensive survey of transformer-based extractive summarization models for researchers and describes the evaluation measures. © 2024 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Enhanced Network Intrusion Detection Using a Hybrid CNN-LSTM Approach on the UNSW-NB15 Dataset 21

Enhanced Network Intrusion Detection Using a Hybrid CNN-LSTM...

引用

21st International Conference on electrical engineering, Computing Science and Automatic Control, CCE 2024

作者： Pear, Zareen Tasnim Kibria, Hafsa Binte Rajshahi University of Engineering and Technology Department of Electrical and Computer Engineering Rajshahi6204 Bangladesh

ISBN: (纸本)9798350377545

In the era of rapid technological advancements, data and information have become significant assets that require continuous protection from unauthorized access and malicious attacks. Network Intrusion Detection Systems (NIDS) can effectively manage the increasing data demands assuring robust security. However, traditional or manual methods are often inadequate to cope with the sophisticated and evolving nature of cyber threats due to their inability to quickly process large volumes of data. Traditional machine learning models are widely used for this purpose. Recently deep learning approaches have become popular in classification tasks showing promising results. This paper introduces a hybrid deep-learning approach for detecting various network intrusions using binary and multi-class (ten classes) classification on the well-known UNSW-NB15 dataset. The proposed hybrid CNN-LSTM model demonstrates significant improvements in intrusion detection, achieving an accuracy of 97.19% for binary classification and 87.70% for multi-class classification, particularly in handling complex attack patterns and reducing false positives. Evaluated using metrics such as precision, recall and F1-score, the model consistently performs robustly across nine distinct attack categories. These results highlight the model's effectiveness in establishing enhanced network security. © 2024 IEEE.

关键词： Network intrusion

来源：评论

学校读者我要写书评

暂无评论

Demand side management with wireless channel impact in IoT-enabled smart grid system

引用

Digital Communications and Networks 2025年第2期11卷 493-504页

作者： Md.Farhad Hossain Kumudu S.Munasinghe Nishant Jagannath Khandakar Tanvir Ahmed Md.Nabid Hasan Ibrahim Elgendi Abbas Jamalipour Department of Electrical and Electronic Engineering Bangladesh University of Engineering and TechnologyDhaka-1205Bangladesh School of Information Technology and Systems University of CanberraACT 2617Australia School of Electrical and Information Engineering The University of SydneyNSW 2006Australia

Demand Side Management(DSM)is a vital issue in smart grids,given the time-varying user demand for electricity and power generation cost over a *** the other hand,wireless communications with ubiquitous connectivity and low latency have emerged as a suitable option for smart *** design of any DSM system using a wireless network must consider the wireless link impairments,which is missing in existing *** this paper,we propose a DSM system using a Real-Time Pricing(RTP)mechanism and a wireless Neighborhood Area Network(NAN)with data transfer uncertainty.A Zigbee-based Internet of Things(IoT)model is considered for the communication infrastructure of the NAN.A sample NAN employing XBee and Raspberry Pi modules is also implemented in real-world settings to evaluate its reliability in transferring smart grid data over a wireless *** proposed DSM system determines the optimal price corresponding to the optimum system welfare based on the two-way wireless communications among users,decision-makers,and energy providers.A novel cost function is adopted to reduce the impact of changes in user numbers on electricity *** results indicate that the proposed system benefits users and energy ***,experimental results demonstrate that the success rate of data transfer significantly varies over the implemented wireless NAN,which can substantially impact the performance of the proposed DSM *** simulations are then carried out to quantify and analyze the impact of wireless communications on the electricity price,user welfare,and provider welfare.

关键词： Smart grid Real time pricing Demand side management Wireless communications Zigbee

来源：评论

学校读者我要写书评

暂无评论

Regression analysis of squeezing-induced hybrid nanofluid flow in Darcy-Forchheimer porous medium

引用

Applied Mathematics and Mechanics(English Edition) 2025年第1期46卷 193-208页

作者： K.MUHAMMAD M.SARFRAZ Department of Humanities and Sciences School of Electrical Engineering and Computer ScienceNational University of Sciences and TechnologyIslamabad 44000Pakistan Department of Mathematics Quaid-i-Azam UniversityIslamabad 44000Pakistan

This article presents a mathematical model addressing a scenario involving a hybrid nanofluid flow between two infinite parallel *** plate remains stationary,while the other moves downward at a squeezing *** space between these plates contains a Darcy-Forchheimer porous medium.A mixture of water-based fluid with gold(Au)and silicon dioxide(Si O2)nanoparticles is *** contrast to the conventional Fourier's heat flux equation,this study employs the Cattaneo-Christov heat flux equation.A uniform magnetic field is applied perpendicular to the flow direction,invoking magnetohydrodynamic(MHD)***,the model accounts for Joule heating,which is the heat generated when an electric current passes through the *** problem is solved via NDSolve in *** and statistical analyses are conducted to provide insights into the behavior of the nanomaterials between the parallel plates with respect to the flow,energy transport,and skin *** findings of this study have potential applications in enhancing cooling systems and optimizing thermal management *** is observed that the squeezing motion generates additional pressure gradients within the fluid,which enhances the flow rate but reduces the frictional ***,the fluid is pushed more vigorously between the plates,increasing the flow *** the fluid experiences higher flow rates due to the increased squeezing effect,it spends less time in the region between the *** thermal relaxation,however,abruptly changes the temperature,leading to a decrease in the temperature fluctuations.

关键词： convective boundary condition Darcy-Forchheimer medium hybrid nanofuid Joule heating magnetohydrodynamic(MHD) numerical solution squeezing flow regression analysis

来源：评论

学校读者我要写书评

暂无评论

Feature Selection-Based Machine Learning Approaches for Detecting Android Malware with Explainable AI 3

Feature Selection-Based Machine Learning Approaches for Dete...

引用

3rd International Conference on Advancement in electrical and Electronic engineering, ICAEEE 2024

作者： Habib, Mahbuba Kibria, Hafsa Binte Rajshahi University of Engineering & Technology Department of Electrical & Computer Engineering Rajshahi6204 Bangladesh

ISBN: (纸本)9798350388282

With the involvement of technology, its negative sides are showing out drastically. As Android is an open-source platform that has nice customization, control and accessibility features cost-effectively, it attracts middle-class device users. As of 2023, Android has a 70.29% global market share. So, its insecurity has a tremendous effect on the mobile industry. This study represents how to efficiently detect Android malware in machine learning with various feature selection techniques like Correlation, Mutual Information, Chi-Square test and Recursive Feature Elimination to serve the optimization of the huge amount of information collected from different apps. Besides, these criteria are tested with different under-sampling and oversampling techniques. This research evaluates some more widely used algorithms: Decision Tree, Random Forest, and XG Boost. Among them, Random Forest surpasses all other algorithms. The output is evaluated with another dataset. Also, Explainable AI is used for the transparency of the models. This implies that Smote NC-based Mutual Information gives the highest outcome(98.5%) with an F1 score of 98% using a 10-fold cross-validation (CV) in Android malware detection. © 2024 IEEE.

关键词： Android malware

来源：评论

学校读者我要写书评

暂无评论

Optimizing UPVC profile production using adaptive neuro-fuzzy inference system

引用

International Journal of Information technology (Singapore) 2024年 1-18页

作者： Naghipour, Avaz Salehpour, Arash Iranag, Behnam Safiri Department of Computer Engineering Faculty of Electrical and Computer Engineering University of Tabriz Tabriz Iran

This paper explores the implementation of an Adaptive Neuro-Fuzzy Inference System to optimize Unplasticized Polyvinyl Chloride profile production. Given the intrinsic complexities of polymer extrusion, such as maintaining consistent quality amidst varying raw material properties and environmental conditions, traditional control methods like Proportional-Integral-Derivative (PID) controllers often need to catch up in adapting to non-linear production environments. This study proposes an ANFIS-based control system that leverages the adaptability of neural networks and the human-like reasoning of fuzzy logic to address these challenges effectively. The system uses real-world data from expert operators to refine its control parameters, ensuring accurate regulation of critical variables such as extrusion temperature, feed material composition, and ambient conditions. By testing the system against traditional control methods, the research demonstrates significant improvements in quality control, efficiency, and waste reduction. The ANFIS system’s adaptability allows it to minimize errors and adjust dynamically to variations in production conditions, outperforming traditional PID systems in maintaining consistent quality. Furthermore, the study outlines the implications of this research for intelligent manufacturing within Industry 4.0 frameworks, highlighting how ANFIS can integrate seamlessly with existing manufacturing systems to optimize production. This paper ultimately contributes to the growing literature advocating for intelligent control systems in polymer extrusion. It underscores their potential to transform manufacturing efficiency and quality in Unplasticized Polyvinyl Chloride production. © Bharati Vidyapeeth's Institute of computer Applications and Management 2024.

关键词： Adaptive neuro-fuzzy inference system (ANFIS) Fuzzy logic controller Manufacturing optimization Neural networks Polymer extrusion process UPVC profile production

来源：评论

学校读者我要写书评

暂无评论

Metaheuristic-Driven Two-Stage Ensemble Deep Learning for Lung/Colon Cancer Classification

引用

computers, Materials & Continua 2024年第9期80卷 3855-3880页

作者： Pouyan Razmjouei Elaheh Moharamkhani Mohamad Hasanvand Maryam Daneshfar Mohammad Shokouhifar Department of Electrical&Computer Engineering Valie-e-Asr University of RafsanjanKerman***Iran International Academic Relations Office(IRO) Computer Science DepartmentUniversity of HalabjaHalabja46018Iraq Department of Computer Engineering University of Mohaghegh ArdabiliArdabil***Iran Department of Biomedical Engineering Islamic Azad University of MashhadMashhad***Iran Department of Electrical&Computer Engineering Shahid Beheshti UniversityTehran***Iran

This study investigates the application of deep learning,ensemble learning,metaheuristic optimization,and image processing techniques for detecting lung and colon cancers,aiming to enhance treatment efficacy and improve survival *** introduce a metaheuristic-driven two-stage ensemble deep learning model for efficient lung/colon cancer *** diagnosis of lung and colon cancers is attempted using several unique indicators by different versions of deep Convolutional Neural Networks(CNNs)in feature extraction and model constructions,and utilizing the power of various Machine Learning(ML)algorithms for final ***,we consider different scenarios consisting of two-class colon cancer,three-class lung cancer,and fiveclass combined lung/colon cancer to conduct feature extraction using four *** extracted features are then integrated to create a comprehensive feature *** the next step,the optimization of the feature selection is conducted using a metaheuristic algorithm based on the Electric Eel Foraging Optimization(EEFO).This optimized feature subset is subsequently employed in various ML algorithms to determine the most effective ones through a rigorous evaluation *** top-performing algorithms are refined using the High-Performance Filter(HPF)and integrated into an ensemble learning framework employing weighted *** findings indicate that the proposed ensemble learning model significantly surpasses existing methods in classification accuracy across all datasets,achieving accuracies of 99.85%for the two-class,98.70%for the three-class,and 98.96%for the five-class datasets.

关键词： Lung cancer colon cancer feature selection electric eel foraging optimization deep learning ensemble learning

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：