检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Cinà, Antonio Emanuele Villani, Francesco Pintor, Maura Schönherr, Lea Biggio, Battista Pelillo, Marcello Department of Computer Science Bioengineering Robotics and Systems University of Genoa Italy Department of Electrical and Electronic Engineering University of Cagliari Italy CISPA Helmholtz Center for Information Security Germany Department of Environmental Sciences Informatics and Statistics Ca’ Foscari University of Venice Italy

Evaluating the adversarial robustness of deep networks to gradient-based attacks is challenging. While most attacks consider 2- and ∞-norm constraints to craft input perturbations, only a few investigate sparse 1- and 0-norm attacks. In particular, 0-norm attacks remain the least studied due to the inherent complexity of optimizing over a non-convex and non-differentiable constraint. However, evaluating adversarial robustness under these attacks could reveal weaknesses otherwise left untested with more conventional 2- and ∞-norm attacks. In this work, we propose a novel 0-norm attack, called σ-zero, which leverages a differentiable approximation of the 0 norm to facilitate gradient-based optimization, and an adaptive projection operator to dynamically adjust the trade-off between loss minimization and perturbation sparsity. Extensive evaluations using MNIST, CIFAR10, and ImageNet datasets, involving robust and non-robust models, show that σ-zero finds minimum 0-norm adversarial examples without requiring any time-consuming hyperparameter tuning, and that it outperforms all competing sparse attacks in terms of success rate, perturbation size, and efficiency. Copyright © 2024, The Authors. All rights reserved.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

NYCTALE: NEURO-EVIDENCE TRANSFORMER FOR ADAPTIVE AND PERSONALIZED LUNG NODULE INVASIVENESS PREDICTION

arXiv

引用

arXiv 2024年

作者： Khademi, Sadaf Oikonomou, Anastasia Plataniotis, Konstantinos N. Mohammadi, Arash Concordia Institute for Information Systems Engineering Concordia University Montreal Canada Department of Medical Imaging Sunnybrook Health Sciences Centre Toronto Canada Department of Electrical and Computer Engineering University of Toronto Toronto Canada

Drawing inspiration from the primate brain’s intriguing evidence accumulation process, and guided by models from cognitive psychology and neuroscience, the paper introduces the NYCTALE framework, a neuro-inspired and evidence accumulation-based Transformer architecture. The proposed neuro-inspired NYCTALE offers a novel pathway in the domain of Personalized Medicine (PM) for lung cancer diagnosis. In nature, Nyctales are small owls known for their nocturnal behavior, hunting primarily during the darkness of night. The NYCTALE operates in a similarly vigilant manner, i.e., processing data in an evidence-based fashion and making predictions dynamically/adaptively. Distinct from conventional Computed Tomography (CT)-based Deep Learning (DL) models, the NYCTALE performs predictions only when sufficient amount of evidence is accumulated. In other words, instead of processing all or a pre-defined subset of CT slices, for each person, slices are provided one at a time. The NYCTALE framework then computes an evidence vector associated with contribution of each new CT image. A decision is made once the total accumulated evidence surpasses a specific threshold. Preliminary experimental analyses conducted using a challenging in-house dataset comprising 114 subjects. The results are noteworthy, suggesting that NYCTALE outperforms the benchmark accuracy even with approximately 60% less training data on this demanding and small dataset. © 2024, CC BY.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

PrivShape: Extracting Shapes in Time Series Under User-Level Local Differential Privacy

PrivShape: Extracting Shapes in Time Series Under User-Level...

引用

International Conference on Data engineering

作者： Yulian Mao Qingqing Ye Haibo Hu Qi Wang Kai Huang Department of Computer Science and Engineering Southern University of Science and Technology Department of Electrical and Electronic Engineering The Hong Kong Polytechnic University Research Institute of Trustworthy Autonomous Systems Southern University of Science and Technology School of Computer Science and Engineering Macau University of Science and Technology

ISBN: (数字)9798350317152

ISBN: (纸本)9798350317169

Time series have numerous applications in finance, healthcare, IoT, and smart city. In many of these applications, time series typically contain personal data, so privacy infringement may occur if they are released directly to the public. Recently, local differential privacy (LDP) has emerged as the state-of-the-art approach to protecting data privacy. However, existing works on LDP-based collections cannot preserve the shape of time series. A recent work, PatternLDP, attempts to address this problem, but it can only protect a finite group of elements in a time series due to ω-event level privacy guarantee. In this paper, we propose PrivShape, a trie-based mechanism under user-level LDP to protect all elements. PrivShape first transforms a time series to reduce its length, and then adopts trie-expansion and two-level refinement to improve utility. By extensive experiments on real-world datasets, we demonstrate that PrivShape outperforms PatternLDP when adapted for offline use, and can effectively extract frequent shapes.

关键词： Privacy Differential privacy Shape Smart cities Frequency-domain analysis Time series analysis Transforms

来源：评论

学校读者我要写书评

暂无评论

A Study of Different Machine Learning Approaches for Relation Discovery Task

A Study of Different Machine Learning Approaches for Relatio...

引用

2025 International Conference on Energy, Power, Environment, Control and Computing, ICEPECC 2025

作者： AL-Aswadi, Fatima N. Za’al Alma’aitah, Wafa Chan, Huah Yong Gan, Keng Hoon Iqbal, Zahid Al-Dhamari, Ahlam Institute of Computer Science and Digital Innovation UCSI University Kuala Lumpur56000 Malaysia Faculty of Computer Science and Engineering Hodeidah University P.O. Box 3114 Hodeidah Yemen Department of Intelligence Systems Faculty of Artificial Intelligence Al-Balqa Applied University Salt19117 Jordan School of Computer Sciences Universiti Sains Malaysia Pulau Pinang Gelugor11800 Malaysia Department of Computer Science Air University Kharian Pakistan Department of Computer Science University of Gujrat Gujrat Pakistan Department of Electrical and Computer Engineering Curtin University Malaysia Sarawak Miri98009 Malaysia

ISBN: (纸本)9781837243235

Relation extraction and classification, collectively referred to as relation discovery, are fundamental tasks in the ontology construction process. In recent years, numerous studies have dedicated their efforts to devising hybrid approaches, supporting inclusive innovation in computational linguistics and ontology engineering. These approaches merge traditional methods with Machine Learning (ML) or Deep Learning (DL) techniques to tackle relation discovery tasks. This research examines the applicability and efficacy of four distinct models— Support Vector Machine (SVM), Naive Bayes (NB), k-Nearest Neighbors (KNN), and Convolutional Neural Network (CNN)—in relation extraction and classification. Two different datasets, ACL RD-TEC 2.0 and ACLRelAcS, are utilized, alongside two distinct text representation methodologies, BOW and TF-IDF-BOW. An in-depth analysis of the results, including the advantages and limitations of each model, is presented in the results and discussion sections. © The Institution of engineering & Technology 2025.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Unveiling a New Vulnerability in Modern Power systems: Leveraging Publicly-Available LMPs for Crafting Cyber-Attacks

Unveiling a New Vulnerability in Modern Power Systems: Lever...

引用

On-Line Conference (ONCON), Industrial Electronics Society Annual

作者： Mohammadmahdi Asghari Amir Ameli Mohsen Ghafouri Mohammad N. Uddin Department of Electrical and Computer Engineering Lakehead University Thunder Bay Canada Concordia Institute for Information Systems Engineering (CIISE) Concordia University Montreal Canada

Modern power systems are increasingly vulnerable to stealthy false data injection attacks (FDIAs) targeting state estimation. Executed without the operator’s knowledge, these attacks can lead to severe consequences, including economic losses, power outages, and full system blackouts. Previously, it was believed that executing destructive FDIAs requires full understanding of the operator’s response to such attacks, which demands in-depth knowledge of power system details, including generation prices and system constraints. Nevertheless, this paper reveals a noteworthy vulnerability wherein an attacker can employ publicly accessible historical locational marginal price (LMP) data and either all or a limited subset of phasor measurement unit (PMU) data to construct a predictive model. This model, utilizing a neural network, takes the loading profile of a section or the entire system as its input to forecast the LMPs. Such a model can be applied in designing high-damage FDIAs. Rigorous testing on the IEEE 30-, IEEE 57-, and IEEE 118- bus test systems demonstrates the model’s ability to predict the LMPs of the system with impressive accuracy and precision. This finding not only validates the model’s efficacy but also underscores the pressing need to address the newfound vulnerability of modern power systems against attackers utilizing public data.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Editorial: Urban multi-energy system networks with high proportion of renewable energy

引用

Frontiers in Energy Research 2025年 13卷

作者： Ding, Xiaoyi Xiang, Yue Morais, Hugo School of Water Resources and Hydroelectric Engineering Xi’an University of Technology Xi’an China Department of Electrical Engineering Sichuan University Chengdu China Institute of Systems and Computer Engineering: Research and Development Higher Technical Institute University of Lisbon Lisbon Lisbon Portugal

Faced with environment pollution and climate change, it is essential to develop advanced low-carbon energy system based on renewable sources. Micro-energy networks with a high proportion of renewable energy are one of the most important methods to achieve the complementary use of renewable energy and fossil energy with high efficiency of energy utilization. Research on urban-level energy networks would be an important step for applying and integrating renewable sources. Faced with high penetration rate of renewable energy, the robust ability and safety operation are challenging from both sides of energy generation and the electrical grid. Meanwhile, the increasing number of electrical and fuel cell powered vehicles put additional difficulties for optimization and scheduling. Therefore, research on urban energy systems with a high proportion of renewable energy is extremely *** topic (“Urban Multi-energy System Networks with High Proportion of Renewable Energy”) aims to look into the key problems existing in the integrated multi-energy network system for the urban level and provide a detailed theoretical basis and experimental data for the design and operation of integrated micro-energy networks at the international academic level. The topic received several submissions within the field of renewable energy systems, including modeling and optimization, grid supply, the interaction between grid and EV/fuel cell cars, and advanced controlling strategy of off-grid/grid-connected microgrid systems, ***-carbon development of energyFeng et al. [1] modeled the natural gas system, power system, and their coupling components, and proposed a low-carbon economic control framework based on MASAC. The objective was to minimize both operational costs and carbon emissions. Simulation results using a nodal system demonstrated that this method significantly reduces costs and emissions compared to traditional approaches. Hu et al. [2] developed a low-carbon scheduling mod

关键词： renewable energy electric vehicle (EV) dynamic control multi-energy system synthetic fuel

来源：评论

学校读者我要写书评

暂无评论

Adaptive Block Elevation Mapping for Large-scale Scene

Adaptive Block Elevation Mapping for Large-scale Scene

引用

2022 IEEE International Conference on robotics and Biomimetics, ROBIO 2022

作者： Zhou, Yang Zhu, Shiqiang Huang, Huang Li, Yuehua Gu, Jason The Research Center for Intelligent Robotics Research Institute of Interdisciplinary Innovation Zhejiang Laboratory Hangzhou311100 China Zhejiang Engineering Research Center for Intelligent Robotics Hangzhou311100 China The Beijing Institute of Control Engineering Beijing100094 China The Beijing Research Institute Zhejiang Laboratory Beijing100094 China The Department of Electrical and Computer Engineering Dalhousie University HalifaxNSB3H 4R2 Canada

ISBN: (纸本)9781665481090

Dense map that contains the surrounding geometry and vision information of a robot is widely used for path planning, navigation, obstacle avoidance and other applications. Considering the performance of the processing unit mounted on the robot is limited, mapping algorithm has to make compromise by sacrificing speed and precision. It will be more challenging when the dense mapping scene is very large because the memory consumption will be greatly increased and the map is difficult to be extended if beyonding the initial map. To suppress the negative impact from the increased map scale, we proposed a novel block mapping approach to generate the dense map in large scale of scene. In this work, the elevation map is selected as the base dense map. The entire elevation map is segmented into numerous block maps of which size is much smaller than that of the entire map. The present moment of lidar and vision measurements are used to generate the local elevation map. The local elevation map is used to update block maps which are adaptively generated along the motion trajectory. A memory-disk interaction mechanism, which the block maps will be loaded to memory or saved to local disk when needed, is introduced. Our block mapping approach is tested on the KITTI datasets, and the results demonstrate that the mapping approach can stably operate in a large scale of scene with a much smaller consumption of memory. © 2022 IEEE.

关键词： Motion planning

来源：评论

学校读者我要写书评

暂无评论

Enhanced Detection of Demagnetization in Direct-Drive Wind Turbine Generators Using eMIST and Machine Learning

Enhanced Detection of Demagnetization in Direct-Drive Wind T...

引用

IEEE Workshop on electrical Machines Design, Control and Diagnosis (WEMDCD)

作者： Qiuyi Chen Panagiotis Tsilifis Jaime Renedo Anglada Department of Mechanical Engineering University of Maryland College Park MD USA Robotics AI & Computer Vision Group GE Vernova Advanced Research Niskayuna NY USA Electrical Systems GE Vernova Advanced Research Niskayuna NY USA

ISBN: (数字)9798331520748

ISBN: (纸本)9798331520755

The rapid growth of offshore wind energy needs robust condition monitoring to ensure the reliability and efficiency of direct-drive permanent magnet synchronous generators (PMSGs). Demagnetization, a critical fault in these generators, can severely impact performance and increase operational costs, especially in harsh marine environments. Traditional diagnostic techniques often struggle to detect subtle magnetic property changes under varying load conditions. This paper presents an innovative framework for demagnetization detection in PMSGs, leveraging the electrical Multi-phase Imbalance Separation Technique (eMIST) combined with machine learning models. Using a dataset from offshore wind turbines, including simulated faulty conditions, the proposed methodology enhances anomaly detection by isolating imbalance signatures and correlating them with key operational parameters. Gaussian Process Regression (GPR) is employed to predict anomalies, offering high accuracy and confidence in fault classification. Results demonstrate the approach's ability to identify early-stage demagnetization, enabling predictive maintenance, reducing downtime, and improving the economic viability of offshore wind farms.

关键词： Costs Accuracy Wind energy Machine learning Gaussian processes Wind farms Generators Wind turbines Demagnetization Anomaly detection

来源：评论

学校读者我要写书评

暂无评论

Slow Inter-area Electro-mechanical Oscillations Revisited: Structural Property of Complex Multi-area Electric Power systems

Slow Inter-area Electro-mechanical Oscillations Revisited: S...

引用

International Conference on Control, Decision and Information Technologies (CoDIT)

作者： Hiya Akhil Gada Marija D. Ilic Laboratory for Information and Decision Systems (LIDS) of Massachusetts Institute of Technology Cambridge MA USA Department of Carnegie Electrical and Computer Engineering (ECE) Mellon University Pittsburgh PA

ISBN: (数字)9798350373974

ISBN: (纸本)9798350373981

This paper introduces a physically-intuitive notion of inter-area dynamics in systems comprising multiple interconnected energy conversion modules. The idea builds on an earlier general approach of setting their structural properties by modeling internal dynamics in stand-alone modules (components, areas) using the fundamental conservation laws between energy stored and generated, and then constraining explicitly their Tellegen’s quantities (power and rate of change of power). In this paper we derive, by following the same principles, a transformed state-space model for a general nonlinear system. Using this model we show the existence of an area-level interaction variable, intVar, whose rate of change depends solely on the area internal power imbalance and is independent of the model complexity used for representing individual module dynamics in the area. Given these structural properties of stand-alone modules, we define in this paper for the first time an inter-area variable as the difference of power wave incident to tie-line from Area I and the power reflected into tie-lie from Area II. Notably, these power waves represent the interaction variables associated with the two respective interconnected areas. We illustrate these notions using a linearized case of two lossless inter-connected areas, and show the existence of a new inter-area mode when the areas get connected. We suggest that lessons learned in this paper open possibilities for computationally-efficient modeling and control of inter-area oscillations, and offer further the basis for modeling and control of dynamics in changing systems comprising faster energy conversion processes.

关键词： Computational modeling Power system dynamics Process control Energy conversion Nonlinear dynamical systems State-space methods Complexity theory Information technology Oscillators

来源：评论

学校读者我要写书评

暂无评论

Absolute Pose from One or Two Scaled and Oriented Features

Absolute Pose from One or Two Scaled and Oriented Features

引用

Conference on computer Vision and Pattern Recognition (CVPR)

作者： Jonathan Ventura Zuzana Kukelova Torsten Sattler Dániel Baráth Department of Computer Science & Software Engineering Cal Poly San Luis Obispo Faculty of Electrical Engineering Visual Recognition Group Czech Technical University in Prague Czech Institute of Informatics Robotics and Cybernetics Czech Technical University in Prague Department of Computer Science Computer Vision and Geometry Group ETH Zürich

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

Keypoints used for image matching often include an estimate of the feature scale and orientation. While recent work has demonstrated the advantages of using feature scales and orientations for relative pose estimation, relatively little work has considered their use for absolute pose estimation. We introduce minimal solutions for absolute pose from two oriented feature correspondences in the general case, or one scaled and oriented correspondence given a known vertical direction. Nowadays, assuming a known direction is not particularly restrictive as modern consumer devices, such as smartphones or drones, are equipped with Inertial Measurement Units (IMU) that provide the gravity direction by default. Compared to traditional absolute pose methods requiring three point correspondences, our solvers need a smaller minimal sample, reducing the cost and complexity of robust estimation. Evaluations on large-scale and public real datasets demonstrate the advantage of our methods for fast and accurate localization in challenging conditions. Code is available at https://***/danini/absolute-pose-from-oriented-and-sealed-features.

关键词： Location awareness Accuracy Simultaneous localization and mapping Measurement units Pose estimation Noise Inertial navigation

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：