检索结果-内蒙古大学图书馆

Digital Sovereignty in the Era of Industry 5.0: Challenges and Opportunities

Procedia computer Science 2025年 254卷 108-117页

作者： Sanjay Misra Kousik Barik Petter Kvalvik Department of Applied Data Science Institute for Energy Technology Halden Norway Department of Computer Sciences Universidad de Alcalá Alcalá de Henares Madrid Spain Digital Systems Institute for Energy Technology Halden Norway

This paper examines the core concept of digital sovereignty within the Industry 5.0 framework, highlighting the importance of sovereign control over data, digital infrastructure, and digital artifacts such as algorithms, models, and visualizations. Trustworthy, autonomous, distributed architectures are essential to building resilience and independence in a multipolar world. Through a comprehensive analysis of existing literature, this paper identifies key challenges and opportunities in achieving digital sovereignty, using a PRISMA chart to select relevant studies from six databases systematically. The article offers strategic insights on extending digital technologies to ensure digital sovereignty while encouraging independent and ethical innovation. By identifying research gaps and proposing future study areas, this work aims to inspire further exploration and development in securing digital sovereignty, supporting resilient digital value chains, and advancing the goals of Industry 5.0.

关键词： Industry 5.0 Digital Sovereignty Privacy Cybersecurity AI tools

来源：评论

学校读者我要写书评

暂无评论

SEMU: Singular Value Decomposition for Efficient Machine Unlearning

arXiv

引用

arXiv 2025年

作者： Sendera, Marcin Struski, Lukasz Ksiazek, Kamil Musiol, Kryspin Tabor, Jacek Rymarczyk, Dawid Faculty of Mathematics and Computer Science Jagiellonian University Poland Institute of Theoretical and Applied Informatics Polish Academy of Sciences Poland Ardigen SA Poland

While the capabilities of generative foundational models have advanced rapidly in recent years, methods to prevent harmful and unsafe behaviors remain underdeveloped. Among the pressing challenges in AI safety, machine unlearning (MU) has become increasingly critical to meet upcoming safety regulations. Most existing MU approaches focus on altering the most significant parameters of the model. However, these methods often require fine-tuning substantial portions of the model, resulting in high computational costs and training instabilities, which are typically mitigated by access to the original training dataset. In this work, we address these limitations by leveraging Singular Value Decomposition (SVD) to create a compact, low-dimensional projection that enables the selective forgetting of specific data points. We propose Singular Value Decomposition for Efficient Machine Unlearning (SEMU), a novel approach designed to optimize MU in two key aspects. First, SEMU minimizes the number of model parameters that need to be modified, effectively removing unwanted knowledge while making only minimal changes to the model’s weights. Second, SEMU eliminates the dependency on the original training dataset, preserving the model’s previously acquired knowledge without additional data requirements. © 2025, CC BY.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Selectively sharing experiences improves multi-agent reinforcement learning 23

Selectively sharing experiences improves multi-agent reinfor...

引用

Proceedings of the 37th International Conference on Neural Information Processing Systems

作者： Matthias Gerstgrasser Tom Danino Sarah Keren School of Engineering and Applied Sciences Harvard University Computer Science Department Stanford University The Taub Faculty of Computer Science Technion - Israel Institute of Technology

We present a novel multi-agent RL approach, Selective Multi-Agent Prioritized Experience Relay, in which agents share with other agents a limited number of transitions they observe during training. The intuition behind this is that even a small number of relevant experiences from other agents could help each agent learn. Unlike many other multi-agent RL algorithms, this approach allows for largely decentralized training, requiring only a limited communication channel between agents. We show that our approach outperforms baseline no-sharing decentralized training and state-of-the art multi-agent RL algorithms. Further, sharing only a small number of highly relevant experiences outperforms sharing all experiences between agents, and the performance uplift from selective experience sharing is robust across a range of hyperparameters and DQN variants.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Optimal Prediction of Heart Disease by Identifying the Type of Chest Pain Using Machine Learning Techniques 4th

Optimal Prediction of Heart Disease by Identifying the Type ...

引用

4th International Conference on Communication and Intelligent Systems, ICCIS 2022

作者： Ahmad, Ghulab Nabi Fatima, Hira Shafiullah Noor, Arshil Institute of Applied Sciences Mangalayatan University Uttar Pradesh Aligarh India Department of Mathematics K.C.T.C College BRA Bihar University Muzaffarpur Raxual India Department of Computer Science Institute of Technology and Management Uttar Pradesh Aligarh India

ISBN: (纸本)9789819920990

Any serious heart ailment is referred to as coronary heart disease. Researchers are concentrating on developing smart systems to identify heart disease precisely based on electronic health data, with the use of machine learning algorithms (MLAs), because they can be fatal. This study introduces ten machine learning techniques (MLTs) for predicting heart diseases using patient data on key healthcare outcomes as: logistic regression classifier (LRC), decision tree classifier (DTC), random forest classifier (RFC), extra tree classifier (EXTC), K-nearest neighbours classifier (KNNC), support vector machine classifier (SVMC), bagging classifier (BC), gradient boosting classifier (GBC), light GBM classifier (LGBMC), Gaussian Naive Bayes classifier (GNBC) on standard datasets from Cleveland, Hungarian, Switzerland, Long Beach VA database, and Statlog Heart Disease of University of California Irvine (UCI) repository to build the prediction models. Data preprocessing and feature selection steps are done before building the models. The models are evaluated based on the accuracy, precision, recall, and F1-score, specificity, MCC, ROC–AUC, and balanced accuracy. The extra tree classifier performed best with 89.10% accuracy, 93.50% precision, 87.80% recall, 90.06% F1-score, 91.10% specificity, 78.00% MCC, 89.44% ROC–AUC, and 89.10% balanced accuracy. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Nearest neighbor search

来源：评论

学校读者我要写书评

暂无评论

Use of IOT in computer Control Using Hand Gesture Using AT Mega 328 over the Cloud

Use of IOT in Computer Control Using Hand Gesture Using AT M...

引用

International Conference on ICT in Business Industry & Government (ICTBIG)

作者： Ishwarlal Rathod Kunal Rathod Neha Gupta Computer Science and Information Technology Symbiosis University of Applied Sciences Indore India Computer Science and Engineering Acropolis Institute of Technology and Research Indore India

In the realm of remote control, my preference is computer manipulation through hand gestures. Recent strides in technology showcase the feasibility of using manual gestures on Arduino platforms, facilitated by cloud computing. A Wi-Fi-based hand gesture computer control system was developed using an ESP-01 Wi-Fi module and an Arduino microcontroller, enabling remote control and monitoring without geographical constraints or time limitations. The ESP8266 Wi-Fi module played a crucial role in establishing wireless communication with an Android application (S Remote) for transmitting commands to regulate domestic appliances. The AT mega 328 AVR series Microcontroller R3, lacking built-in Wi-Fi, required the ESP-01 Wi-Fi module for wireless communication and programming. The Arduino IDE facilitated programming, incorporating ESP-01 and LCD components. Hand gesture-controlled robots and PC/laptop hand gesture control systems, both utilizing Arduino technology, exemplify similar applications. The Human Machine Interface (HMI) concept involves physical components and software enabling communication between users and machines. HMI devices, including LED indicators and touch screens, traditionally use keyboards and mice. This research implemented a hand gesture controller using Arduino technology, specifically for web browser features, video playback, and VLC media player audio adjustments. The ESP-01 Wi-Fi module, configured via AT instructions through a UART serial connection, played a key role in integrating a hand gesture system for effective computer control. Users can perform actions like playing or pausing videos and adjusting audio amplitude within the VLC media player solely through manual gestures.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Consensus MPC-Based hierarchical control of Islanded AC Microgrid

Consensus MPC-Based hierarchical control of Islanded AC Micr...

引用

International Conference on Communication, Computing and Digital Systems (C-CODE)

作者： Zawar Khan Muhammad Aamir Electrical & Computer Engineering Department Pak-Austria Fachhochschule: Institute of Applied Sciences and Technology Haripur Pakistan

In this paper, the consensus model predictive control (MPC)-based hierarchical control scheme is presented for the primary and secondary layers of an islanded AC microgrid (MG). The secondary control is designed on the tracker consensus method, in this way a distributed controller in collaboration with the finite time observer will set the referenced values of voltage magnitude and frequency by significantly speeding up the method of synchronous convergence. Furthermore, the primary control has been implemented for each DG separately and comprises Finite Control Set-Model Predictive control (FCS-MPC), and droops control. FCS-MPC regulates PCC reference voltages for each DG unit. Droop Control regulates power exchange within and among DG units. In the proposed hierarchical control approach, information sharing is accomplished by a neighbouring node creating a sparse communication network while fully eliminating the requirement for a centralized controller. The controller effectiveness have been validated against unbalanced load, transients, and load plug-and-play capabilities. The simulation result shows that MG operates effectively using FCS-MPC for parallel inverters of DG units, and the consensus algorithm based distributed secondary controller in MG effectively delivers the anticipated power, voltage, and frequency.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Metadata Enhanced Security Watermarks for Sensor Data Protection

Metadata Enhanced Security Watermarks for Sensor Data Protec...

引用

2023 IEEE applied Sensing Conference, APSCON 2023

作者： Kondapuram, Akash Reddy Treytl, Albert Ruotsalainen, Henri Sauter, Thilo Danube University Krems Dept. for Integrated Sensor Systems Wiener Neustadt Austria St. Pölten University of Applied Sciences Institute of It Security Research St. Pölten Austria Institute of Computer Technology Tu Wien Vienna Austria

ISBN: (纸本)9781665461634

The Internet of Things (IoT) and Home Automation Systems are rapidly evolving areas of technology. Since these systems are commonly used with lightweight sensor nodes which have limited computational and communications capabilities, low-cost measures for integrity protection and authenticity verification are required to protect sensor data. Dynamic water-marking of sensor data is an approach to detect various kinds of sensor attacks such as hardware manipulation, network replay attacks or sensor data modification attacks both at the sensor source node and also on the connecting network. We propose a metadata enhanced digital watermark, which bits are embedded in the (least significant bits) LSB's of the transmitted sensor data to reach zero payload size increase. © 2023 IEEE.

关键词： Metadata

来源：评论

学校读者我要写书评

暂无评论

Explaining BERT model decisions for near-duplicate news article detection based on named entity recognition

Explaining BERT model decisions for near-duplicate news arti...

引用

International Conference on Semantic Computing

作者： Anne Stockem Novo Fatih Gedikli Institute of Computer Science University of Applied Sciences Ruhr West Mülheim an der Ruhr Germany

While it is a straight-forward task to detect lexically identical documents, it is challenging to do this on documents that are semantically similar. This work deals with the detection of near-duplicate news articles on various websites. The detection of documents of related content is especially interesting for search engines and recommender systems. Search engines aim at eliminating duplicates to prevent the user from seeing the same document in different versions. Recommender systems filter out items that are too similar to those the user already knows to tackle the problem of overspecialization. A near-duplicate is in some cases a mere paraphrasing of a previous article, in other cases it contains corrections or added content as a follow-up article. In this work, we investigate, among other things, the extent to which named entities such as people, places, and organisations that appear in the article text can be used to detect near-duplicates. Evaluating a fine-tuned BERT model on common named entities in an article pair, performance measures greater than 97 % are achieved. The SHAP library is then used to get insight into the model decisions and feature importance, which is the individual words in a text document. Overall, the results show that named entities are very suitable for the detection of near-duplicates. Furthermore, the explanations are in agreement with an intuitive human interpretation.

关键词： Deep learning Measurement Bit error rate Semantics Search engines Data models Libraries

来源：评论

学校读者我要写书评

暂无评论

Supporting the Requirements-Based Selection of Suitable Communication Protocols in Smart Grids

Supporting the Requirements-Based Selection of Suitable Comm...

引用

2022 IEEE PES Innovative Smart Grid Technologies Conference Europe, ISGT-Europe 2022

作者： Hanna, Sebastian Rohjans, Sebastian Heeren, Philipp Rolink, Johannes Institute for Information Technology Division of Energy Offis Oldenburg Germany Jade University of Applied Sciences Civil Engineering Geoinformation and Health Technology Oldenburg Germany University of Applied Sciences Emden/Leer Electrical Engineering and Computer Science Emden Germany

ISBN: (数字)9781665480321

ISBN: (纸本)9781665480321

The implementation of the smart grid and the related use cases requires the interaction of a wide variety of systems from many manufacturers. The selection of suitable communication protocols is a key factor in achieving the necessary interoperability and avoiding vendor lock-ins. Considering the large number of possible protocols, this selection process is becoming increasingly difficult and time-consuming. To simplify and accelerate this process, the use of a morphological box is presented. The partial aspects defined in the morphological box are chosen in such a way that they can describe both the requirements of a use case and the possibilities of the protocol under assessment. Applying the box eventually results in reusable profiles that allow the comparison of use cases and protocols in a quick and simple way. The utilization is demonstrated by evaluating the MQTT protocol for the realization of a given use case. © 2022 IEEE.

关键词： Smart power grids

来源：评论

学校读者我要写书评

暂无评论

Real-Time Human Fall Detection using a Lightweight Pose Estimation Technique

arXiv

引用

arXiv 2024年

作者： Alam, Ekram Sufian, Abu Dutta, Paramartha Leo, Marco Department of Computer Science Gour Mahavidyalaya West Bengal India Department of Computer Science University of Gour Banga India Department of Computer and System Sciences Visva-Bharati University India National Research Council of Italy Institute of Applied Sciences and Intelligent Systems Lecce73100 Italy

The elderly population is increasing rapidly around the world. There are no enough caretakers for them. Use of AI-based in-home medical care systems is gaining momentum due to this. Human fall detection is one of the most important tasks of medical care system for the aged people. Human fall is a common problem among elderly people. Detection of a fall and providing medical help as early as possible is very important to reduce any further complexity. The chances of death and other medical complications can be reduced by detecting and providing medical help as early as possible after the fall. There are many state-of-the-art fall detection techniques available these days, but the majority of them need very high computing power. In this paper, we proposed a lightweight and fast human fall detection system using pose estimation. We used 'Movenet' for human joins key-points extraction. Our proposed method can work in real-time on any low-computing device with any basic camera. All computation can be processed locally, so there is no problem of privacy of the subject. We used two datasets 'GMDCSA' and 'URFD' for the experiment. We got the sensitivity value of 0.9375 and 0.9167 for the dataset 'GMDCSA' and 'URFD' respectively. The source code and the dataset GMDCSA of our work are available online to access. © 2024, CC BY-NC-ND.

关键词： Fall detection

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：