检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Schuchardt, Jan Stoian, Mihail Kosmala, Arthur Günnemann, Stephan Dept. of Computer Science Munich Data Science Institute Technical University of Munich Germany Dept. of Engineering University of Technology Nuremberg Germany

Amplification by subsampling is one of the main primitives in machine learning with differential privacy (DP): Training a model on random batches instead of complete datasets results in stronger privacy. This is traditionally formalized via mechanism-agnostic subsampling guarantees that express the privacy parameters of a subsampled mechanism as a function of the original mechanism’s privacy parameters. We propose the first general framework for deriving mechanism-specific guarantees, which leverage additional information beyond these parameters to more tightly characterize the subsampled mechanism’s privacy. Such guarantees are of particular importance for privacy accounting, i.e., tracking privacy over multiple iterations. Overall, our framework based on conditional optimal transport lets us derive existing and novel guarantees for approximate DP, accounting with Rényi DP, and accounting with dominating pairs in a unified, principled manner. As an application, we analyze how subsampling affects the privacy of groups of multiple users. Our tight mechanism-specific bounds outperform tight mechanism-agnostic bounds and classic group privacy results. Copyright © 2024, The Authors. All rights reserved.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

Predictive Analysis of Consumer Behavior Using Supervised Learning Techniques 2

Predictive Analysis of Consumer Behavior Using Supervised Le...

引用

2nd International Conference Computational and Characterization Techniques in Engineering and sciences, IC3TES 2024

作者： Rekha, K. Sashi Amutha, T. Usharani, R. Pushparani, S. Sivagami, V.M. Saveetha Institute of Medical and Technical Sciences Simats Engineering Department of Computer Science and Engineering Tamil Nadu Chennai India Care College of Engineering Department of Artificial Intelligence & Data Science Tamil Nadu Trichy India Srm Institute of Science and Technology Faculty of Engineering and Technology Department of Computational Intelligence School of Computing Chennai India Meenakshi College of Engineering Department of Computer Science and Engineering Tamil Nadu Chennai India Sri Venkateswara College of Engineering Department of Information Technology Tamil Nadu Chennai India

ISBN: (纸本)9798350364699

The study presented here consists of an entire framework for prediction of consumers behavior through the machine training of supervised learning methods. The strategy involves the following: data collection, preprocessing, exploration, feature selection, model selection, training, evaluation, interpretation and deployment. The systematic process is tailored to clear the pathway for combining data about customer taste, sophisticated habits, and purchasing trends. Primary components include data preprocessing, training of machine learning models, feature selection, and ethical issues regarding data privacy and transparency. The outcomes of the implementation are a high accuracy in consumers clustering, efficient classification of customers' behavior, as well as unveiling a bunch of crucial predictors that affect the customers mood and behavior. As a result, the report stresses the necessity of integrated monitoring and redirection strategies to realize the potential of the created models. The idea of predictive analysis as an evolving tool that enables marketers to tailor their campaigns more closely, and improve the overall customer experience therefore also receives attention. The proposed system methodology constitutes a solid process framework for business defined to aim at using predictive analytics to gain competitive edge in the current dynamic marketplace. © 2024 IEEE.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

AI-Enhanced Public Safety Lighting for Smart Sustainable Cities Using Cloud Computing

AI-Enhanced Public Safety Lighting for Smart Sustainable Cit...

引用

2025 International Conference on Multi-Agent Systems for Collaborative Intelligence, ICMSCI 2025

作者： Tidke, Bharat Pragadeeswaran, S. Lakshmi, R. Poornima Bhuvaneswari, G. Manikandan, G. Murugan, S. University Nagpur Campus Pune India School of Computing Vel Tech Rangarajan Dr. Sagunthala R&d Institute of Science and Technology Department of Computer Science and Engineering Tamil Nadu Chennai India Chennai Institute of Technology Department of Computer Science and Engineering Chennai India Saveetha Engineering college Department of Computer Science and Engineering Thandalam Tamil Nadu Chennai India Rmk Engineering College Department of Artificial Intelligence and Data Science RSM Nagar Kavaraipettai Tamil Nadu Chennai India Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Saveetha University Department of Biomedical Engineering Tamil Nadu Chennai India

ISBN: (纸本)9798331509828

This paper presents a novel method that combines cloud computing infrastructure and Reinforcement Learning (RL) algorithms to improve public safety lighting in smart cities. By dynamically adjusting lighting levels in response to real-time sensor inputs, the proposed system aims to maximize visibility while minimizing energy consumption. The cloud-based framework's RL agents continuously interact with the environment and human activity patterns to learn the best lighting policies. The system utilizes data from multiple sensors, including ambient light and motion detectors, to adjust real-time to daylight levels, weather, and pedestrian traffic variations. By integrating RL, the lighting system can adapt to changing urban dynamics and safety requirements over time, all while reducing energy waste. Compared to conventional static lighting systems, case studies and simulations show how the system can significantly improve visibility and response times while enhancing safety in various urban settings. RL algorithms can be easily deployed and managed throughout extensive smart city networks due to cloud computing network scalability and adaptability. This study advances sustainable urban infrastructure by showcasing the effectiveness of RL-based artificial intelligence (AI) solutions in public safety lighting optimization, thereby laying the groundwork for future smarter and safer cities. © 2025 IEEE.

关键词： Sustainable city

来源：评论

学校读者我要写书评

暂无评论

Adversarial Training for Graph Neural networks: Pitfalls, Solutions, and New Directions

arXiv

引用

arXiv 2023年

作者： Gosch, Lukas Geisler, Simon Sturm, Daniel Charpentier, Bertrand Zügner, Daniel Günnemann, Stephan Department of Computer Science Munich Data Science Institute Technical University of Munich Germany Microsoft Research United States

Despite its success in the image domain, adversarial training did not (yet) stand out as an effective defense for Graph Neural networks (GNNs) against graph structure perturbations. In the pursuit of fixing adversarial training (1) we show and overcome fundamental theoretical as well as practical limitations of the adopted graph learning setting in prior work;(2) we reveal that flexible GNNs based on learnable graph diffusion are able to adjust to adversarial perturbations, while the learned message passing scheme is naturally interpretable;(3) we introduce the first attack for structure perturbations that, while targeting multiple nodes at once, is capable of handling global (graph-level) as well as local (node-level) constraints. Including these contributions, we demonstrate that adversarial training is a state-of-the-art defense against adversarial structure perturbations.1 © 2023, CC BY.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Transforming Shopping with IoT: Exploring the Intelligent Way to Design the Smart Trolley using IoT Principles

Transforming Shopping with IoT: Exploring the Intelligent Wa...

引用

2024 IEEE International Conference on Big data and Machine Learning, ICBDML 2024

作者： Jayaprakash, M. Hency Juliet, A. Jaya Mabel Rani, A. Velusudha, N.T. Lakshmisridevi, S. Samuthira Pandi, V. R.M.K. Engineering College Department of Information Technology Tamil Nadu Kavaraipettai India Saveetha University Saveetha Institute of Medical and Technical Sciences Department of Data Analytics Tamil Nadu Chennai India Saveetha Institute of Medical and Technical Sciences Department of Computer Science and Engineering Tamil Nadu Chennai India Prince Shri Venkateshwara Padmavathy Engineering College Department of Electronics and Communication Engineering Chennai India BS Abdur Rahman Crescent Institute of Science and Technology Department of Computer Science and Engineering Tamil Nadu Chennai India Integrated Technology Chennai Institute of Technology Centre for Advanced Wireless Tamil Nadu Chennai India

ISBN: (纸本)9798350374100

The interconnection of commonplace items is revolutionizing people's daily lives, thanks to the Internet of Things (IoT). A smart shopping system, consisting of interconnected products, might one day be a reality in supermarkets. In this Internet of Things system, products may be individually identified using low-cost RFID tags, which, when added to an RFID-enabled smart shopping cart, can be read automatically. This means shoppers won't have to endure the checkout line at all - billing can be handled right in the cart. The integration of RFID-equipped smart shelving into this system allows for the monitoring of inventory and, perhaps, the update of a central server. The suggested shopping system, which is dubbed IoT Assisted Smart Trolley (IoTAST), is cross-validated with the standard Smart Trolley (ST) model to clearly evaluate its performance. Since an RFID reader can automatically read all objects instead of a labourer physically scanning them, inventory management becomes considerably easier with this sort of technology. We construct a working system to test its functioning, determine the design criteria of a sophisticated purchasing system, and then create a reliable communication mechanism that renders the system feasible. This study validates the viability of the system. This is the first smart shopping system that takes security into account that we are aware of. © 2024 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Hierarchical Randomized Smoothing

arXiv

引用

arXiv 2023年

作者： Scholten, Yan Schuchardt, Jan Bojchevski, Aleksandar Günnemann, Stephan Dept. of Computer Science Munich Data Science Institute Technical University of Munich Germany University of Cologne Germany

Real-world data is complex and often consists of objects that can be decomposed into multiple entities (e.g. images into pixels, graphs into interconnected nodes). Randomized smoothing is a powerful framework for making models provably robust against small changes to their inputs – by guaranteeing robustness of the majority vote when randomly adding noise before classification. Yet, certifying robustness on such complex data via randomized smoothing is challenging when adversaries do not arbitrarily perturb entire objects (e.g. images) but only a subset of their entities (e.g. pixels). As a solution, we introduce hierarchical randomized smoothing: We partially smooth objects by adding random noise only on a randomly selected subset of their entities. By adding noise in a more targeted manner than existing methods we obtain stronger robustness guarantees while maintaining high accuracy. We initialize hierarchical smoothing using different noising distributions, yielding novel robustness certificates for discrete and continuous domains. We experimentally demonstrate the importance of hierarchical smoothing in image and node classification, where it yields superior robustness-accuracy trade-offs. Overall, hierarchical smoothing is an important contribution towards models that are both – certifiably robust to perturbations and accurate.1 © 2023, CC BY-SA.

关键词： Pixels

来源：评论

学校读者我要写书评

暂无评论

Smart Personal Protective Equipment in Ambulance Services with IoT Integration for Safety 2

Smart Personal Protective Equipment in Ambulance Services wi...

引用

2nd International Conference on Smart Technologies for Smart Nation, SmartTechCon 2023

作者： Senkamalavalli, R. Merlin Livingston, L.M. Guruva Reddy, Elangovan Basavaraddi, Chethan Chandra S. Latha, N. Department of Computer Science and Engineering Kammavari Sangham Institute of Technology Karnataka Bangalore India Department of Electronics and Communication Engineering Jeppiaar Institute of Technology Tamil Nadu Chennai India Department of Artificial Intelligence and Data Science Koneru Lakshmaiah Education Foundation Andhra Pradesh Vaddeswaram India Department of Computer Science and Engineering Kalpataru Institute of Technology Tiptur Karnataka Tumkur India Department of Computer Science and Engineering Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Saveetha University Tamil Nadu Chennai India

ISBN: (纸本)9798350305418

Internet of Things (IoT) technology has improved ambulance safety and efficiency. In ambulance services, Smart PPE (Personal Protective Equipment) improves patient and healthcare worker safety via IoT connectivity. In crises, ambulances are vital. However, healthcare professionals risk contagious infections and physical damage. Though vital, traditional PPE lacks real-time monitoring and communication, leaving healthcare workers exposed. These issues are addressed with Smart PPE using IoT. This novel invention adds sensors and communication devices to paramedic and EMT PPE. These sensors measure air quality, temperature, and the wearer's heart and body temperature. A centralized system can remotely monitor healthcare personnel's safety and well-being using real-time data. Smart PPE has two-way communicators and emergency alert systems. Emergency coordination is improved by ambulance crews and dispatch centers communicating seamlessly. The Smart PPE can automatically inform the team and dispatch in urgent situations, increasing response times and patient outcomes. IoT enables predictive maintenance and data analytics. Trends and dangers may be identified from gathered data, allowing proactive risk mitigation. Smart PPE can measure use and remind users to maintain and replace equipment, assuring dependability when required most. By continually monitoring conditions, allowing real-time communication, and enabling data-driven decision-making, this unique technology improves ambulance services and reduces dangers to devoted personnel. © 2023 IEEE.

关键词： Ambulance Services Efficiency Improvement Emergency Healthcare Real-time Alerts Safety Enhancement

来源：评论

学校读者我要写书评

暂无评论

LLM-Safety Evaluations Lack Robustness

arXiv

引用

arXiv 2025年

作者： Beyer, Tim Xhonneux, Sophie Geisler, Simon Gidel, Gauthier Schwinn, Leo Günnemann, Stephan Department of Computer Science Munich Data Science Institute Technical University of Munich Germany Mila & Université de Montréal Canada Canada AI CIFAR Canada

In this paper, we argue that current safety alignment research efforts for large language models are hindered by many intertwined sources of noise, such as small datasets, methodological inconsistencies, and unreliable evaluation setups. This can, at times, make it impossible to evaluate and compare attacks and defenses fairly, thereby slowing progress. We systematically analyze the LLM safety evaluation pipeline, covering dataset curation, optimization strategies for automated red-teaming, response generation, and response evaluation using LLM judges. At each stage, we identify key issues and highlight their practical impact. We also propose a set of guidelines for reducing noise and bias in evaluations of future attack and defense papers. Lastly, we offer an opposing perspective, highlighting practical reasons for existing limitations. We believe that addressing the outlined problems in future research will improve the field’s ability to generate easily comparable results and make measurable progress. © 2025, CC BY.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Fusion of DRL and CNN for Effective Face Recognition 1

引用

8th International Conference on Information System Design and Intelligent Applications, ISDIA 2024

作者： Kumar, Ankit Singh, Sushil Kumar Ravikumar, R.N. Khanna, Ashish Brahma, Biswajit Dr. APJ Abdul Kalam Technical University Lucknow India Department of Computer Engineering Marwadi University Rajkot India Computer Science and Engineering Department Maharaja Agrasen Institute of Technology Delhi India Center for Global Health Research Saveetha Medical College Saveetha Institute of Medical and Technical Sciences Chennai India Senior Data and Visualization Engineer McKesson Corporation USA 32559 Lake Bridgeport St FremontCA94555 United States

ISBN: (数字)9789819765812

ISBN: (纸本)9789819765805

Digital Authentication combined with IOT networks is one of the most innovative practices in innovations. IoT devices such as surveillance cameras have acquired real-time video data in the natural environment. Facial Biometrics for System authentication is considered to be a sophisticated technology used in various devices. The correct processing of digital information in a natural environment is one of the challenging tasks. Noise effect inducts random variations on the subject in the natural environment that may hinder the recognition efficiency of the model. The proposed system recognizes facial biometric features under the natural environment that may contain non-symmetrical randomness of variations. The system's objective is to mitigate the false recognition rates under adverse environmental conditions. It combines digital information with a person’s real-time environment. The detection of the facial region of interest form frames in a real-time dataset has been taken through the Viola-Jones algorithm. Feature extrication using video frames has been developed through Deep Reinforcement Learning (DRL) algorithm, which aims to generate binary trees containing feature vectors. Further, the system uses a convolutional neural network (CNN) model to establish the correlation of feature vectors belonging to the facial identity. The model's objective is to retain high recognition of facial biometric trait feature units under various randomness in a natural environment. The proposed system is also tested under various attacks to test the robustness of the proposed model. The model can secure an average accuracy of 98.85%. Digital forensics, crime investigation, online attendance, etc., are the few applications of Facial data authentication from dynamic input. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Sustainable Smart Agriculture Advisor through Artificial Intelligence Approach

Sustainable Smart Agriculture Advisor through Artificial Int...

引用

2024 International Conference on Distributed Systems, computer networks and Cybersecurity, ICDSCNC 2024

作者： Hariharan, Shanmugasundaram Srivani, Pogula Gogula, Sreenivasulu Venkateswaran, N. Kukreja, Vinay Jose, P. Vardhaman College of Engineering Dept. of Artificial Intelligence and Data Science Hyderabad India Cmr Technical Campus Department of H&s Hyderabad India Vardhaman College of Engineering Department of Computer Science and Engineering Hyderabad India Jyothishmathi Institute of Technology & Science Nustulapur Department of Computer Science and Engineering Karimnagar India Centre for Research Impact & Outcome Chitkara University Rajpura India Vel Tech Rangarajan Dr. Sagunthala R&d Institute of Science and Technology Department of Cse Chennai India

ISBN: (纸本)9798350375442

Agriculture, constituting the backbone of India's economy and employing 70% of its population, faces critical challenges such as water scarcity, soil degradation, and unpredictable crop yields. Traditional farming methods lack efficiency, contributing to environmental impact and resource inefficiency. To address these issues on smart crop cultivation with artificial intelligence enabled crop advisor sustainable agriculture generation integrates machine learning and data analytic to offer real-time insights and recommendations. This advanced agricultural approach aims to optimize resource utilization, minimize environmental impact, and adapt to changing climatic conditions. By empowering farmers with actionable insights, Smart Crop seeks to revolutionize agriculture, promoting sustainable practices for increased resource efficiency and enhanced resilience against challenges like limited precision farming, climate variability, and yield maximization. This innovative solution endeavors to bridge the gap between traditional farming methods and modern challenges, fostering a more sustainable and productive future for Indian agriculture. The methods analyzed leads to promising results as compared with well known metrics. © 2024 IEEE.

关键词： Smart agriculture

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：