检索结果-内蒙古大学图书馆

3rd IEEE International Conference on Automation and Computation, AUTOCOM 2025

作者： Jayaprakash, B. Premkumar, J. Das, Neeraj Punia, Ankit Nayak, Subrat Kumar Ojha, Varun Karnataka Bangalore India Sathyabama Institute of Science and Technology Department of Biomedical Engineering Tamil Nadu Chennai India Maharishi School of Engineering & Technology Maharishi University of Information Technology Uttar Pradesh India Chitkara University Centre of Research Impact and Outcome Punjab Rajpura140417 India Department of Computer Science and Engineering Odisha Bhubaneswar India Chitkara University Chitkara Centre for Research and Development Himachal Pradesh Baddi174103 India

ISBN: (纸本)9798331542375

Wireless body area networks are an attractive technology for measuring health data in medical applications. Several challenges exist in designing and implementing such reliable, energy-efficient communication protocols for these networks. This new strategy solution strategy is based on a hybrid communication protocol that simultaneously obtains the positive aspects of centralized and distributed structures. By centralizing network maintenance and control functions such as health monitoring of nodes and transmission control, the proposed protocol supports reliable data transmission with minimal energy consumption in WBANs. This reduces the overhead and collisions, giving the network better reliability. Unlike other protocols mentioned where the data flow from the sensor node to the central node is at least distributed, in our case, a few nodes send their data at a time to the Central Node. It enables low-power communication since nodes only need to wake up during their allotted transmission time slot. Through simulations, we demonstrate that the new protocol surpasses existing protocols in reliability and energy consumption performance. It can also help to minimize the energy consumption of WBAN nodes, which makes it an appropriate candidate for long-time medical monitoring applications. All in all, this work helps to pave the way for the efficient use of WBANs in a wide range of medical applications that require robustness and low-energy communications. © 2025 IEEE.

关键词： Signaling

来源：评论

学校读者我要写书评

暂无评论

Attackers Can Do Better: Over- and Understated Factors of Model Stealing Attacks

Attackers Can Do Better: Over- and Understated Factors of Mo...

引用

Secure and Trustworthy Machine Learning (SaTML), IEEE Conference on

作者： Daryna Oliynyk Rudolf Mayer Andreas Rauber CDL AsTra Faculty of Computer Science University of Vienna Vienna Austria SBA Research Vienna Austria Faculty of Informatics TU Wien Institute of Information Systems Engineering Vienna Austria

ISBN: (数字)9798331517113

ISBN: (纸本)9798331517120

Machine learning (ML) models were shown to be vulnerable to model stealing attacks, which lead to intellectual property infringement. Among other attack methods, substitute model training is an all-encompassing attack applicable to any machine learning model whose behaviour can be approximated from input-output queries. Whereas previous works mainly focused on improving the performance of substitute models by, e.g. developing a new substitute training method, there have been only limited ablation studies that try to understand the impact the strength of an attacker has on the substitute model's performance. As a result, different authors came to diverse, sometimes contradicting, conclusions. In this work, we exhaustively examine the ambivalent influence of different factors resulting from varying the attacker's capabilities and knowledge on a substitute training attack. Our findings suggest that some of the factors that have been considered important in the past are, in fact, not that influential; instead, we discover new correlations between attack conditions and success rate. In particular, we demonstrate that better-performing target models enable higher-fidelity attacks and explain the intuition behind this phenomenon. Further, we propose to shift the focus from the complexity of target models toward the complexity of their learning tasks. Therefore, for the substitute model, rather than aiming for a higher architecture complexity, we suggest focusing on getting data of higher complexity and an appropriate architecture. Finally, we demonstrate that even in the most limited data-free scenario, there is no need to overcompensate weak knowledge with unrealistic capabilities in the form of millions of queries. Our results often exceed or match the performance of previous attacks that assume a stronger attacker, suggesting that these stronger attacks are likely endangering a model owner's intellectual property to a significantly higher degree than shown until now.

关键词： Training Correlation Query processing Focusing Intellectual property Data collection Data models Adversarial machine learning Complexity theory Tuning

来源：评论

学校读者我要写书评

暂无评论

Attackers Can Do Better: Over- and Understated Factors of Model Stealing Attacks

arXiv

引用

arXiv 2025年

作者： Oliynyk, Daryna Mayer, Rudolf Rauber, Andreas CDL AsTra Faculty of Computer Science University of Vienna Vienna Austria SBA Research Vienna Austria Institute of Information Systems Engineering Faculty of Informatics TU Wien Vienna Austria

Machine learning (ML) models were shown to be vulnerable to model stealing attacks, which lead to intellectual property infringement. Among other attack methods, substitute model training is an all-encompassing attack applicable to any machine learning model whose behaviour can be approximated from input-output queries. Whereas previous works mainly focused on improving the performance of substitute models by, e.g. developing a new substitute training method, there have been only limited ablation studies that try to understand the impact the strength of an attacker has on the substitute model’s performance. As a result, different authors came to diverse, sometimes contradicting, conclusions. In this work, we exhaustively examine the ambivalent influence of different factors resulting from varying the attacker’s capabilities and knowledge on a substitute training attack. Our findings suggest that some of the factors that have been considered important in the past are, in fact, not that influential;instead, we discover new correlations between attack conditions and success rate. In particular, we demonstrate that better-performing target models enable higher-fidelity attacks and explain the intuition behind this phenomenon. Further, we propose to shift the focus from the complexity of target models toward the complexity of their learning tasks. Therefore, for the substitute model, rather than aiming for a higher architecture complexity, we suggest focusing on getting data of higher complexity and an appropriate architecture. Finally, we demonstrate that even in the most limited data-free scenario, there is no need to overcompensate weak knowledge with unrealistic capabilities in the form of millions of queries. Our results often exceed or match the performance of previous attacks that assume a stronger attacker, suggesting that these stronger attacks are likely endangering a model owner’s intellectual property to a significantly higher degree than shown until now. Cop

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Iterative Motion Planning in Multi-agent systems with Opportunistic Communication under Disturbance

arXiv

引用

arXiv 2025年

作者： Thelasingha, Neelanga Julius, Agung Humann, James Dotterweich, James Dept. of Electrical Computer and Systems Engineering Rensselaer Polytechnic Institute NY United States James Humann and James Dotterweich are with DEVCOM Army Research Laboratory MD United States

— In complex multi-agent systems involving heterogeneous teams, uncertainty arises from numerous sources like environmental disturbances, model inaccuracies, and changing tasks. This causes planned trajectories to become infeasible, requiring replanning. Further, different communication architectures used in multi-agent systems give rise to asymmetric knowledge of planned trajectories across the agents. In such systems, replanning must be done in a communication-aware fashion. This paper establishes the conditions for synchronization and feasibility in epistemic planning scenarios introduced by opportunistic communication architectures. We also establish conditions on task satisfaction based on quantified recoverability of disturbances in an iterative planning scheme. We further validate these theoretical results experimentally in a UAV–UGV task assignment problem. © 2025, CC BY.

关键词：

来源：评论

学校读者我要写书评

暂无评论

引用

Expert systems with Applications 2025年 282卷

作者： Wang, Dan Xia, Yuze Pedrycz, Witold Li, Zhiwu Yu, Zhenhua College of Computer Science and Technology Xi'an University of Science and Technology Xi'an710054 China Department of Electrical and Computer Engineering University of Alberta EdmontonABT6G 2R3 Canada Systems Research Institute Polish Academy of Sciences Warsaw00-901 Poland Research Center of Performance and Productivity Analysis Istinye University Istanbul Turkey Institute of Systems Engineering Macau University of Science and Technology Macau900000 China School of Electro-Mechanical Engineering Xidian University Xi'an710071 China

Deep learning models have achieved remarkable success in the field of computer vision. However, it is difficult to understand and explain the complex internal mechanisms of deep learning models. Existing explanation methods often suffer from issues such as random noises and feature redundancy, which may mislead users and obscure the practical decision-making process of the models. To address these issues and improve the interpretation quality, we propose a novel method named Feature Similarity Group-Class Activation Mapping (FSG-CAM), which aims to enhance the clarity and quality of model's interpretation and to improve the computational efficiency by reducing noise and redundancy. FSG-CAM introduces leveraging feature similarity grouping to reduce noise and redundancy, thereby achieving clearer and more efficient explanation that align closely with human vision. We first use the gradient of target class computed to the model's output to guide the initial feature map and calculates the similarity between different feature channels. Then, feature channels with similar activation distributions are grouped together, and the sub-channels in each group are stacked to obtain a new feature channel. Next, the mixture masks are formed using normalization and binarization techniques, and the final saliency map is obtained by a linear weighted combination. We conduct extensive experiments based on ImageNet and MS COCO datasets for a collection of qualitative and quantitative analysis on several metrics such as Average Drop/Increase, Insertion/Deletion Curves, noise assessment, and ablation experiments. The results demonstrate that FSG-CAM indeed improves the interpretation quality and produces more consistent explanations with human vision. © 2025 Elsevier Ltd

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Ultra-broadband external-cavity quantum cascade laser synchronization for mode-hop free tuning 24

Ultra-broadband external-cavity quantum cascade laser synchr...

引用

Novel In-Plane Semiconductor Lasers XXIV 2025

作者： Kosan, Nicholas Meyer, Zane Nowicki, Keith Silver, Joel Peterson, Kristen Reyes-Newell, Adriana Applegate, Jeff Rafkin, Scot Wysocki, Gerard Department of Electrical and. Computer Engineering Princeton University PrincetonNJ08544 United States Planetary Science Directorate Southwest Research Institute BoulderCO80302 United States Southwest Sciences Inc. Santa FeNM87505 United States BAE Systems BoulderCO80301 United States Northern Arizona University United States

ISBN: (纸本)9781510685185

The LASVEGAS (Laser Absorption Spectrometer for Volatile and Evolved GASes) instrument is a prototype ultrabroadband external-cavity (EC) quantum cascade laser (QCL) -based spectrometer under development for potential extraterrestrial mission, where simultaneous broad spectral coverage and fine resolution are highly desirable. This is accomplished with an array of EC-QCLs with unique spectral coverage, allowing efficient scaling of the spectral emission range over hundreds of cm-1. To date, the tuning range of the instrument exceeds 600 cm-1 within a span bounded approximately by 944 cm-1 and 1930 cm-1. To access local mode hop free tuning, we leverage a simulation of the EC geometry in order to synchronize the parameters of the EC. While specific mode hop free scans are limited by the dynamic range of the laser drive current and are limited to the scale of 1.5 cm-1, mode hop free sections of adjacent scans can be stitched together. The simulation is used to coordinate the position of the stages in the EC, controlling the EC "air" length and the grating angle. © 2025 SPIE.

关键词： Quantum cascade lasers

来源：评论

学校读者我要写书评

暂无评论

Deep Learning Approaches for Battery Capacity and State of Charge Estimation with the NASA B0005 Dataset

引用

computers, Materials and Continua 2025年第3期83卷 4795-4813页

作者： Zeyang Zhou Zachary James Ryan Utkarsh Sharma Tran Tien Anh Shashi Mehrotra Angelo Greco Jason West Mukesh Prasad School of Computer Science Faculty of Engineering and Information Technology University of Technology Sydney Ultimo NSW 2007 Australia School of Data Science and Engineering Indian Institute of Science Education and Research Bhopal 462066 India Research Institute of Marine Systems Engineering Seoul National University Seoul 08826 Republic of Korea Department of Computer Science and Engineering Faculty of Engineering and Technology SRM Institute of Science and Technology Delhi NCR Campus Ghaziabad 201204 India Battery System Modelling Department CLARIOS VARTA Hannover GmbH Hanover 30419 Germany Science and Innovation Group Bureau of Meteorology Brisbane 4000 Australia

Accurate capacity and State of Charge (SOC) estimation are crucial for ensuring the safety and longevity of lithium-ion batteries in electric vehicles. This study examines ten machine learning architectures, Including Deep Belief Network (DBN), Bidirectional Recurrent Neural Network (BiDirRNN), Gated Recurrent Unit (GRU), and others using the NASA B0005 dataset of 591,458 instances. Results indicate that DBN excels in capacity estimation, achieving orders-of-magnitude lower error values and explaining over 99.97% of the predicted variable’s variance. When computational efficiency is paramount, the Deep Neural Network (DNN) offers a strong alternative, delivering near-competitive accuracy with significantly reduced prediction times. The GRU achieves the best overall performance for SOC estimation, attaining an R 2 of 0.9999, while the BiDirRNN provides a marginally lower error at a slightly higher computational speed. In contrast, Convolutional Neural Networks (CNN) and Radial Basis Function Networks (RBFN) exhibit relatively high error rates, making them less viable for real-world battery management. Analyses of error distributions reveal that the top-performing models cluster most predictions within tight bounds, limiting the risk of overcharging or deep discharging. These findings highlight the trade-off between accuracy and computational overhead, offering valuable guidance for battery management system (BMS) designers seeking optimal performance under constrained resources. Future work may further explore advanced data augmentation and domain adaptation techniques to enhance these models’ robustness in diverse operating conditions.

关键词： Battery capacity estimation state of charge deep learning prediction efficiency energy storage systems

来源：评论

学校读者我要写书评

暂无评论

Empowering Adaptive Endogenous Security Trend Prediction Detection for IoT Sensor Nodes

Empowering Adaptive Endogenous Security Trend Prediction Det...

引用

Advances in Electrical, Electronics and Computing Technology (EECT), International Conference on

作者： Lihu Zhou Xiuying Dong Enli Zhang Ting Wang Xiao Zhang Chong Zhang School of Computer Science and Software Engineering Southwest Petroleum University Chengdu China Exploration and Development Research Institute of PetroChina Southwest Oil & Gasfield Company Chengdu China School of Computer Science Chengdu University of Information Technology Chengdu China

ISBN: (数字)9798331541545

ISBN: (纸本)9798331541552

Massively deployed IoT devices require a lightweight design to reduce costs. However, this architecture inherently limits their security, increasing the risk of data breaches and tampering on a large scale. In this paper, we propose CGAD, a model that leverages one-dimensional convolution (Conv1D), gated recurrent units (GRU), and the attention mechanism to analyze multi-sensor time series data at the data gateway from sensing terminals. To achieve this, we generate time series predictions and detect anomalies by comparing predicted values with real data using a threshold-based approach, thereby enhancing data security. To address the labeling problem of massive data, we employ unsupervised learning and design a dynamic sliding window threshold to improve the judgment of time series with varying characteristics. We also used a number of datasets, such as the KW51 railroad bridge dataset and the self-picked temperature and humidity dataset, to evaluate our system. The results demonstrate that our CGAD model achieves recall rates of 98 % and 99 % for anomaly detection on the two datasets, respectively. Additionally, the precision and F1 scores surpass other methods, ensuring cost-effectiveness for massive terminals and addressing the security issues of sensitive data. These advantages contribute to realizing the Internet of Everything.

关键词： Adaptation models Data security Time series analysis Humidity Predictive models Logic gates Data models Sensors Internet of Things Anomaly detection

来源：评论

学校读者我要写书评

暂无评论

A Novel Approach to Smart Grid Energy Management Using the Osprey Optimization Algorithm 4

A Novel Approach to Smart Grid Energy Management Using the O...

引用

4th International Conference on Sentiment Analysis and Deep Learning, ICSADL 2025

作者： Rudrapogu, Nissisolomon Hemalatha, N. Warkad, Sanjay B Jayanthi, A. Vinoth, K. Jayaram, B. Lecturer in Eee Nissisolomon Rudrapogu Mbts Government Polytechnic College Department of Electrical and Electronics Engineering Section Andhra Pradesh Guntur522005 India Bharath Institute of Higher Education and Research Department of Electrical and Electronics Engineering Tamil Nadu Chennai600073 India P. R. Pote Patil College of Engineering & Management Department of Electrical Engineering Maharashtra Amravati444602 India Jeppiaar Institute of Technology Department of Physics Tamil Nadu Sriperumbudur631502 India Vel Tech Rangarajan Dr. Sagunthala R&d Institute of Science and Technology Department of Electrical and Electronics Engineering Tamil Nadu Chennai600062 India Rmk Engineering College Department of Computer Science and Business Systems Tamil Nadu Thiruvallur601206 India

ISBN: (纸本)9798331523923

Smart Grid Energy Management (SGEM) denotes the function of optimizing energy flow in the grid while enhancing grid efficiency and stability. One major issue is that increased integration of Renewable Energy Sources (RESs) may add a layer of complexity to EM, which may, in turn, reduce the reliability and stability of the grid. Keeping such challenges in mind, this paper proposes the Osprey Optimization Algorithm (OOA) on its part to enhance SGEM through various facets of load balancing and energy distribution. The objective of this method is to minimize the Mean Absolute Percentage Error (MAPE), enhancing the precision of energy predictions and optimizing grid performance. The OOA optimizes energy distribution in micro grids, improving efficiency and reducing power losses. The performance of the suggested OOA method is assessed by contrasting it with other optimization approaches that are currently in use, including Particle Swarm Optimization (PSO), Multi Objective Genetic Algorithm (MOGA), Sparrow Search Algorithm (SSA), and Multi Objective Wind Driven Optimization (MOWDO). The OOA method is implemented on the MATLAB platform. The outcomes show that the proposed OOA method provides superior performance in minimizing MAPE, achieving the lowest MAPE value of 0.28%, indicating its enhanced accuracy in energy demand forecasting and optimization. © 2025 IEEE.

关键词： Clean energy

来源：评论

学校读者我要写书评

暂无评论

C2IDTL: Novel Click to Image Conversion Approach for Deep Transfer Learning in Click Fraud Detection on Digital Platforms

引用

IEEE Access 2025年 13卷 90293-90315页

作者： Fernando, Channa Walgampaya, Chamila Alawatugoda, Janaka University of Peradeniya Computer Center Faculty of Engineering Peradeniya Sri Lanka University of Peradeniya Faculty of Engineering Department of Engineering Mathematics Peradeniya Sri Lanka Rabdan Academy Research and Innovations Centers Division Abu Dhabi United Arab Emirates Griffith University Institute for Integrated and Intelligent Systems NathanQLD Australia

The global retail industry has witnessed a massive expansion of e-commerce businesses in the last decade by transforming physical stores into digitalised entities. One of the biggest threats toe-commerce is click fraud, which generates fraudulent clicks without genuine intention on online advertisements using automated software programs or organised human activities. This has become a major threat to the online advertising eco systems, causing heavy losses for leading tech companies like Google, Facebook, Meta, and Yahoo, which rely on pay per click (PPC) models. To overcome click fraud, we propose a novel click to image conversion technique to generate synthetic images from clicks and perform click fraud detection on those generated images directly with Xception, Inception, VGG, ResNet and DenseNet DTL models (C2IDTL). This novel approach embedded the feature importance and spatial structure into synthetic images to enable the identification of complex pattern recognition via DTL models. We evaluated the performance of C2IDTL on three publicly available datasets, namely FDMA BuzzCity 2012, Talking Data click dataset and Sweden Ad click datasets and compared the performances with the conventional state-of-the-art models. C2IDTL enhances the fraud detection by exceeding 91.18% accuracy, 90.90% F1-score and 96.45% AUC based on the K-fold cross validation performed on all three datasets. Consequently, these results validate the robustness and the ability to generalise across diverse click datasets evaluated in prior scientific literature. Despite DTL’s association with the visual domain, this study challenges its traditional usage and opens avenues for future studies to optimise performance and robustness. © 2013 IEEE.

关键词： Marketplaces

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：