检索结果-内蒙古大学图书馆

NWSTAN: a lightweight dynamic spatial–temporal attention network for traffic prediction

Neural Computing and Applications 2024年 1-16页

作者： Sun, Jingru Zhang, Yao Qiu, Ziyu Cheng, Qixuan Xiao, Zhu Chongqing Research Institute of Hunan University Changsha China College of Computer Science and Electronic Engineering Hunan University Changsha China

Advanced deep learning technology has promoted the development of high-accuracy traffic prediction algorithms. However, existing algorithms struggle to balance model size and predictive accuracy, lack the ability to model the spatial correlation effectively, and require many parameters, resulting in a significantly larger model size. To overcome these challenges, this paper introduces a new solution, the lightweight Nodewalk Spatial–Temporal Attention Network (NWSTAN). NWSTAN uses an improved lightweight nodewalk algorithm based on Node2vec to model the correlations between sensors on different roads dynamically. Additionally, it employs a spatiotemporal feature extraction module that leverages the attention mechanism to extract relevant features from the traffic flow data. Experiments conducted on two real-world datasets show that NWSTAN reduced the mean absolute error (MAE) by 3-5% compared to the baseline model while significantly reducing the number of parameters (by at least 27%). This makes NWSTAN easier to integrate into existing edge computing devices, broadening its potential for real-world applications. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

CAML: Cryptographic-Based Cloud Security for Healthcare Data with Machine Learning Technique

引用

Recent Patents on engineering 2025年第2期19卷 26-40页

作者： M.H, Chaithra Vagdevi, S. Department of Computer Science and Engineering REVA University & Research Scholar Affiliated to Visvesvaraya Technological University Karnataka Belagavi India Department of AI & ML City Engineering College Karnataka Bangalore India

Background: The Cloud model is one of the most realistic frameworks with a vast range of social networking interactions. In medical data, security is a major constraint as it incorporates information about the patients. The cloud environment subjected to mobility and openness is exposed to security issues and limits authorization levels for data transmission. Objective: This patent paper aims to propose a security model for attack prevention within the healthcare environment. Methods: The proposed Cryptographic Attribute-based Machine Learning (CAML) scheme incorporates three stages. Initially, the homomorphic encryption escrow is performed for secure data transmission in the cloud. Secondly, the information of the users is evaluated based on the consideration of users' authorization. The authorization process for the users is carried out with the attribute-based ECC technique. Finally, the ML model with the classifier is applied for the detection and classification of attacks in the medical network. Results: The detected attack is computed and processed with the CNN model. Simulation analysis is performed for the proposed CAML with conventional ANN, CNN, and RNN models. The simulation analysis of proposed CAML achieves a higher accuracy of 0.96 while conventional SVM, RF, and DT achieve an accuracy of 0.82, 0.89 and 0.93, respectively. Conclusion: With the analysis, it is concluded that the proposed CAML model achieves higher classification accuracy for attack detection and prevention in the cloud computing environment. © 2025 Bentham Science Publishers.

关键词： Cloud security

来源：评论

学校读者我要写书评

暂无评论

Intelligent QV2X Routing for Traffic Management in Consumer IoV Using STGNN and Reinforcement Learning

引用

IEEE Transactions on Consumer Electronics 2025年第1期71卷 1288-1297页

作者： Jamal, Syed Saqib Muhammad, Afaq Song, Wang-Cheol Jeju National University Department of Electronic Engineering Jeju Korea Republic of College of Computing and Mathematics Computer Engineering Department and Interdisciplinary Research Center for Intelligent Secure Systems Dhahran Saudi Arabia

In the Consumer Internet of Vehicles (CIoV), reliable and timely data communication is essential for enhancing driver experience and safety. This paper introduces an innovative QV2X routing strategy that uses Spatio-Temporal Graph Neural Networks (STGNN) and Q-learning to optimize packet traffic in CIoV. By predicting network conditions and adapting to real-time data flows, our approach directly addresses consumer needs for efficient data transmission, reduced communication delays, and improved infotainment access. The integration of predictive models with adaptive learning mechanisms not only optimizes packet delivery but also minimizes latency and packet loss, critical for consumer applications like real-time navigation assistance and hazard warnings. Our key contribution is a dynamic packet traffic management system designed for consumer use, enhancing network reliability and efficiency for everyday vehicle users. Experimental results validate that our model surpasses existing benchmarks by improving packet delivery ratios by up to 15% and reducing end-to-end delays by up to 20% in urban traffic scenarios. This advancement demonstrates our strategy's effectiveness in enriching consumer experiences and safety in vehicular communications. © 1975-2011 IEEE.

关键词： Packet loss

来源：评论

学校读者我要写书评

暂无评论

Decentralized Game-based Auction Algorithm for Scheduling Robotic Chargers to Service EVs with Uncertain Demands

IEEE Transactions on Intelligent Vehicles

引用

IEEE Transactions on Intelligent Vehicles 2024年 1-14页

作者： Fang, Qiuyang Zhang, Jianlei Wang, Chen Xie, Guangming Zhang, Chunyan Department of Automation College of Computer and Control Engineering Nankai University Tianjin China National Engineering Research Center of Software Engineering Peking University Beijing China State Key Laboratory for Turbulence and Complex Systems Intelligent Biomimetic Design Lab College of Engineering Peking University Beijing China

In this study, we introduce a novel auction-based algorithm modeled as a decentralized coalition formation game, designed for the complex requirements of large-scale multi-robot task allocation under uncertain demand. This context is particularly illustrative in scenarios where robots are tasked to charge electric vehicles. The algorithm begins by partitioning a composite task sequence into distinct subsets based on spatial similarity principles. Subsequently, we employ a coalition formation game paradigm to coordinate the assembly of robots into cooperative coalitions focused on these distinct subsets. To mitigate the impact of unpredictable task demands on allocations, our approach utilizes the conditional value-at-risk to assess the risk associated with task execution, along with computing the potential revenue of the coalition with an emphasis on risk-related outcomes. Additionally, integrating consensus auctions into the coalition formation framework allows our approach to accommodate assignments for individual robot-task pairings, thus preserving the stability of individual robotic decision autonomy within the coalition structure and assignment distribution. Simulative analyses on a prototypical parking facility layout confirm that our algorithm achieves Nash equilibrium within the coalition structure in polynomial time and demonstrates significant scalability. Compared to competing algorithms, our proposal exhibits superior performance in resilience, task execution efficiency, and reduced overall task completion times. The results demonstrate that our approach is an effective strategy for solving the scheduling challenges encountered by multi-robot systems operating in complex environments. IEEE

关键词： Polynomial approximation

来源：评论

学校读者我要写书评

暂无评论

Learning continuous network emerging dynamics from scarce observations via data-adaptive stochastic processes

引用

Science China(Information Sciences) 2024年第12期67卷 240-255页

作者： Jiaxu CUI Qipeng WANG Bingyi SUN Jiming LIU Bo YANG College of Computer Science and Technology Jilin University Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University Public Computer Education and Research Center Jilin University Department of Computer Science Hong Kong Baptist University

Learning network dynamics from the empirical structure and spatio-temporal observation data is crucial to revealing the interaction mechanisms of complex networks in a wide range of domains. However,most existing methods only aim at learning network dynamic behaviors generated by a specific ordinary differential equation instance, resulting in ineffectiveness for new ones, and generally require dense *** observed data, especially from network emerging dynamics, are usually difficult to obtain, which brings trouble to model learning. Therefore, learning accurate network dynamics with sparse, irregularly-sampled,partial, and noisy observations remains a fundamental challenge. We introduce a new concept of the stochastic skeleton and its neural implementation, i.e., neural ODE processes for network dynamics(NDP4ND), a new class of stochastic processes governed by stochastic data-adaptive network dynamics, to overcome the challenge and learn continuous network dynamics from scarce observations. Intensive experiments conducted on various network dynamics in ecological population evolution, phototaxis movement, brain activity, epidemic spreading, and real-world empirical systems, demonstrate that the proposed method has excellent data adaptability and computational efficiency, and can adapt to unseen network emerging dynamics, producing accurate interpolation and extrapolation with reducing the ratio of required observation data to only about 6% and improving the learning speed for new dynamics by three orders of magnitude.

关键词： complex networks network dynamics emerging spatio-temporal dynamics neural processes

来源：评论

学校读者我要写书评

暂无评论

Enhanced Yoga Posture Detection 15

Enhanced Yoga Posture Detection

引用

15th International Conference on Advances in Computing, Control, and Telecommunication Technologies, ACT 2024

作者： Firke, Jay Bhujbal, Vaishnavi Bodke, Mrudula Warke, Pranav Pimpri Chinchwad College of Engineering and Research Computer Department Pune India

ISBN: (纸本)9798331300579

Human pose detection is a rapidly growing area in computer vision, which has various applications such as action recognition, sports analysis, and human-computer interaction. Yoga, a holistic practice that promotes physical and mental well-being, has gained widespread popularity in recent years. With the advent of technology, there is an increasing demand for innovations that can enhance the yoga experience. The objective is to develop an automated system that can detect and recognize yoga poses in real-time using a live camera. This research tackles the challenging task of real-time yoga pose detection through a live camera feed, aiming to facilitate human-computer interaction and advance the field of computer vision. Yoga poses are complex and highly variable, making their automated detection a noteworthy endeavor. This research aims to develop an iOS application for detecting yoga postures that leverages the capabilities of AV Foundation, Combine, Vision, and Core ML frameworks to construct an advanced system. Its applications extend beyond yoga, encompassing fitness monitoring, rehabilitation, and wellness tracking, further emphasizing its significance in modern society. © Grenze Scientific Society, 2024.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Sarcasm Detection and Classification Using Deep Learning Model 7th

Sarcasm Detection and Classification Using Deep Learning Mod...

引用

7th International Conference on Soft Computing: Theories and Applications, SoCTA 2023

作者： Rajani, B. Saxena, Sameer Suresh Kumar, B. Narang, Geetika AMITY Institute of Information Technology Amity University Rajasthan Jaipur India School of Computer Science and Engineering Sanjay Ghodawat University Kolhapur India Trinity College of Engineering and Research Maharashtra Puna India

ISBN: (纸本)9789819720880

The Twitter and Facebook kind of web platforms have gained significant traction as prominent mediums for individuals to document and articulate their emotions, viewpoints, and evaluations. The use of appropriate extraction methodologies, such as sentiment analysis, renders this data valuable across several domains. Sarcasm detection is a methodology used to examine individuals’ cognitive and affective states, with the aim of classifying. There are several methods by which individuals might articulate their sentiments. These attitudes are occasionally attended with sarcasm, particularly when expressing strong emotions. Sarcasm has often understood as a kind of communication in which a good statement is made with an underlying negative purpose. The majority of existing research endeavors see these responsibilities as separate entities. Until far, the majority of methods for sentiment and sarcasm classification have been focused on treating them as separate and independent text categorization tasks. Nowadays, there has been notable growth in the field of study using deep learning methods, resulting in substantial improvements in the performance of standalone classifiers. One of the primary challenges encountered by these methodologies is their inability to accurately categorize sarcastic statements as negative. Given this consideration, we assert that possessing the ability to identify sarcasm will contribute to the enhancement of sentiment classification, and conversely, proficiency in sentiment classification will aid in the recognition of sarcasm. Our research has shown a positive correlation between these two activities. This study presents a system that employs a deep neural network and multi-task learning to effectively predict the connection between different tasks, with the goal of attracting the overall concert of sentiment analysis. The approach presented in this study demonstrates superior performance compared to the previous methods, with a notable margin o

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Sentimental Analysis on Zomato Restaurant Reviews using Bi-LSTM

Sentimental Analysis on Zomato Restaurant Reviews using Bi-L...

引用

2023 International Conference on Integrated Intelligence and Communication Systems, ICIICS 2023

作者： Chaudhari, Deepti Angadi, Sanjeevkumar Sable, Saili Patil, Uma Chaudhari, Dipamala Jadhav, Kavita Nutan College of Engineering and Research Computer Science and Engineering Pune India

ISBN: (纸本)9798350315455

Sentiment analysis is a vital aspect of understanding public opinion and sentiment towards products and services. This paper presents a sentiment analysis work focused on Zomato restaurant reviews in Bangalore, aiming to classify restaurants into positive, negative, or neutral sentiment categories based on customer reviews. Bi-LSTM and Bi-GRU models are employed to capture contextual information in the sequential data of reviews. Additionally, sentiment analysis techniques, including Word2Vec, VADER, and Sentiment Intensity Analyzer, are integrated to enhance the sentiment classification process. Through rigorous experimentation, the performance of these models and techniques is evaluated. The proposed models demonstrate promising accuracy rates in sentiment classification. By enhancing and expanding the sentiment analysis framework, this paper contributes to a deeper understanding of public sentiment towards Zomato restaurants in Bangalore. The insights derived from this study can facilitate informed decision-making for both restaurant owners and customers, ultimately improving the dining experience and customer satisfaction with accuracy of 98.6%. © 2023 IEEE.

关键词： Bi-GRU Bi-LSTM Restaurant Reviews Sentiment Analysis Text Classification Word2Vec

来源：评论

学校读者我要写书评

暂无评论

A Survey on various Machine Learning approach to predict Health Insurance Cost 1

A Survey on various Machine Learning approach to predict Hea...

引用

1st International Conference on Cognitive Computing and engineering Education, ICCCEE 2023

作者： Bodke, Minal Koparde, Shweta Amrutkar, Rohit Chikurdekar, Abhishek Salunke, Atharv Waghmare, Abhijeet Pimpri Chinchwad College of Engineering and Research Department of Computer Engineering India

ISBN: (纸本)9798350332803

Insurance is a policy that diminish or eradicate the expenditure loss appear due to various risk. Different factors may have an impact on the Insurance cost. Machine learning is a field of study that gives computers potential to adapt without human interference. Machine learning allow insurance companies to acknowledge the customers in a better way and create an insurance cover that victual to their needs and profile. In this paper, the propose system will examine the individuals health details to forecast the health insurance amount prediction and exhibit the insurance plans. Four regression models such as Linear Regression, Support Vector Regressor, Random Forest Regressor and Gradient Boosting have been implemented and their performance is measured using Mean Absolute Error, Root Mean Absolute Error and Coefficient of determination. Based on the analysis, it is found that the Gradient Boosting performs better than the other regression algorithms. © 2023 IEEE.

关键词： Health insurance

来源：评论

学校读者我要写书评

暂无评论

FARMSUPPLY: Food Supply Chain Management using Blockchain Technology 3

FARMSUPPLY: Food Supply Chain Management using Blockchain Te...

引用

3rd IEEE International Conference on Technology, engineering, Management for Societal Impact using Marketing, Entrepreneurship and Talent, TEMSMET 2023

作者： Balekundri, Omkar Tigadi, Raghav Jayakkanavar, Abhilasha Jain College of Engineering and Research Computer Science and Engineering Belgaum India

ISBN: (纸本)9798350397819

The Distributed Ledger Technology (DLT) that underpins various crypto currencies may have a profound impact on the global economy. The Supply Chain Management (SCM) is one of the areas where the DLT is currently being used. By using the Distributed Ledger Technology, blockchain has the potential to vastly enhance different supply chains by facilitating faster and more cost-effective delivery of products. The main benefits of using blockchain in SCM is that it provides an enhancement technique for products traceability, improves coordination among the producer, manufacturer, distributor, retailer and consumer. Also, facilities easier access for funding. In the recent years, the Food Supply Chain (FSC) has been using the blockchain to utilize the benefits of SCM and provide better results for the Food Supply Chain. In this paper, a model is proposed, FARMSUPPLY, which uses blockchain in a Food Supply Chain to provide a better view of farmers, products and retailers. In this model, a Ethereum blockchain method and smart contract for the verification and validation of various attributes at each stage of the food supply chain is presented. © 2023 IEEE.

关键词： Smart contract

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：