检索结果-内蒙古大学图书馆

A Vertical Federated Multi-View Fuzzy Clustering Method for Incomplete data

IEEE Transactions on Fuzzy Systems 2025年第5期33卷 1510-1524页

作者： Li, Yan Hu, Xingchen Yu, Shengju Ding, Weiping Pedrycz, Witold Kiat, Yeo Chai Liu, Zhong National University of Defense Technology Laboratory for Big Data and Decision Hunan Changsha410073 China National University of Defense Technology School of Computer Hunan Changsha410073 China Nantong University School of Artificial Intelligence and Computer Science Nantong226019 China University of Alberta Department of Electrical and Computer Engineering EdmontonABT6G 2R3 Canada Polish Academy of Sciences Systems Research Institute Warsaw00-901 Poland Istinye University Faculty of Engineering and Natural Sciences Department of Computer Engineering Turkey Nanyang Technological University College of Computing and Data Science Singapore639798 Singapore

Multi-view fuzzy clustering (MVFC) has gained widespread adoption owing to its inherent flexibility in handling ambiguous data. The proliferation of privatization devices has driven the emergence of new challenge in MVFC researches. Federated learning, a technique that can jointly train without directly using raw data, has gain significant attention in decentralized MVFC. However, their applicability depends on the assumptions of data integrity and independence between different views. In fact, while within distributed environments, data typically exhibits two challenging problems: (1) multiple views within a single client;(2) incomplete data. Existing methods exhibit limitations in effectively addressing these challenges. Hence, in this study, we aim at achieving the effective clustering for incomplete data by a novel vertical federated MVFC framework. Specifically, a unified clustering framework is designed to capture both local client learning and global server training. For the local client learning, the data reconstruction strategy and prototype alignment strategy are introduced to ensure the preservation of data structure and refinement of clustering relationships, which mitigates the impact of incomplete data. Meanwhile, the global training process implements aggregation based on client-specific information. The whole process is realized based on the unified fuzzy clustering framework, promoting collaborative learning between client-specific and server information. Theoretical analyses and extensive experiments are carefully conducted to validate the effectiveness and efficiency of the proposed method from multiple perspectives. © 1993-2012 IEEE.

关键词： data integrity

来源：评论

学校读者我要写书评

暂无评论

Towards understanding and prediction of corrosion degradation of organic coatings under tropical marine atmospheric environment via a data-driven approach

引用

International Journal of Minerals,Metallurgy and Materials 2025年第5期32卷 1151-1161页

作者： Shaopeng Liu Lingwei Ma Jinke Wang Yiran Li Haiyan Gong Haitao Ren Xiaogang Li Dawei Zhang Beijing Advanced Innovation Center for Materials Genome Engineering Institute for Advanced Materials and TechnologyUniversity of Science and Technology BeijingBeijing 100083China National Materials Corrosion and Protection Data Center University of Science and Technology BeijingBeijing 100083China Institute of Materials Intelligent Technology Liaoning Academy of MaterialsShenyang 110004China Luoyang Ship Material Research Institute Sanya 572032China

The corrosion degradation of organic coatings in tropical marine atmospheric environments results in substantial economic losses across various *** complexity of a dynamic environment,combined with high costs,extended experimental periods,and limited data,places a limit on the comprehension of this *** study addresses this challenge by investigating the corrosion de-gradation of damaged organic coatings in a tropical marine environment using an atmospheric corrosion monitoring sensor and a random forest(RF)*** damage simulation,a polyurethane coating applied to a Fe/graphite corrosion sensor was intentionally scratched and exposed to the marine atmosphere for over one *** correlation analysis was performed for the collection and filtering of en-vironmental and corrosion current *** to the RF model,the following specific conditions contributed to accelerated degrada-tion:relative humidity(RH)above 80%and temperatures below 22.5℃,with the risk increasing significantly when RH exceeded 90%.High RH and temperature exhibited a cumulative effect on coating degradation.A high risk of corrosion occurred in the *** RF model was also used to predict the coating degradation process using environmental data as input parameters,with the accuracy show-ing improvement when the duration of influential environmental ranges was considered.

关键词： organic coating degradation atmospheric corrosion machine learning exposure test random forest coating sensor

来源：评论

学校读者我要写书评

暂无评论

research and Application on the Improvement of Ice Melting Efficiency of Transmission Lines 19th

Research and Application on the Improvement of Ice Melting E...

引用

19th Annual Conference of China Electrotechnical Society, ACCES 2024

作者： Song, Deng Yang, Qi Zeng, Huarong Zhong, Yao Huang, Huan Xie, Quan Meihua Mountain Freezing Climate Disaster and Equipment Safety Observation and Research Station of Guizhou Province Guiyang550002 China Guizhou Power Grid Limited Liability Company Power Science Research Institute Guiyang550002 China School of Big Data and Information Engineering Guizhou University Guiyang550025 China

ISBN: (纸本)9789819613823

Since the large-scale ice disaster occurred in 2008, DC ice melting technology has gradually become a common configuration for substations in medium and heavy ice-covered areas of China, which has become the most effective measure for transmission lines to resist large-scale ice-covered disasters in winter. However, it has been found that many communication links and low ice melting wiring efficiency exist in DC ice melting operation process, causing the transmission line to take too long from outage to resumption. Aiming at the existing problems, the time-consuming characteristics of various wiring forms of ice-melting transmission lines that have been carried out in the past of Guizhou Power Grid are obtained: the dispatcher command model is inefficient in melting ice due to too many communication links. In order to improve work efficiency, this paper proposes an ice melting work-group model that gives transmission line management authority to the person in charge of on-site ice melting work. The advantages of the ice melting work-group mode in the efficiency of ice melting on transmission lines are verified by comparing the two modes in terms of time consumption in engineering application. Finally, for the long time-consuming operation process of Ice-melting wiring operations, a one-key sequence control is proposed, the transformation requirements for substation are given and the efficiency of the method is verified, which has guiding significance for the further improvement of deicing efficiency in the future. © Beijing Paike Culture Commu. Co., Ltd. 2025.

关键词： Efficiency

来源：评论

学校读者我要写书评

暂无评论

data-Driven Stop-Loss Hedging Strategies for NSE Using Machine Learning Techniques

Data-Driven Stop-Loss Hedging Strategies for NSE Using Machi...

引用

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

作者： Nagaraj, P. Prabhu, M. Chaithanya Visist, T. Janani, D. Hitesh, S. Saranath, P. Kalasalingam Academy of Research and Education Department of Computer Science and Engineering Krishnankoil Virudhunagar India Indian Institute of Technology Madras Department of Data Science and Artificial Intelligence Chennai600036 India

ISBN: (纸本)9798331509828

The financial markets are inherently volatile, and investors constantly seek ways to mitigate risks while maximizing returns. This work aims to develop a stop-loss hedging strategy for a portfolio of stocks listed on the National Stock Exchange (Nse) to minimize losses and maximize returns. This strategy utilizes real-time market data to adjust stop-load levels dynamically by predicting future stock. This will protect against market downwards. We used long short-term memory (LSTM) neural networks to predict the next stock value. This AI-driven model allows the system to predict the next stock by using real-time market data and adjust the stoploss level accordingly. In addition, this system integrates a hedging strategy. It determines whether to hedge based on the predicted stock price compared to the current price. If the predicted price is less than the current price a hedging strategy like short selling or buying put option is suggested. This automated solution is useful for retail investors. This strategy will protect against losses and returns maximum profits. When the market is upside the stoploss position also increases and will make maximum profits. If the market is downward it will protest from losses by putting a stoploss level and it will return minimum losses. Here we use different algorithms in the work like decision tree, random forest, LSTM, and SVM. Based on the all algorithms we got the highest accuracy for the decision tree and the accuracy is 99.54%. © 2025 IEEE.

关键词： data Analysis data science Decision Tree Economic Growth Hedging Strategy Long Term Short Memory NSE Random Forest Stocks Stop - Loss SVM

来源：评论

学校读者我要写书评

暂无评论

MASS FLOW DETECTION TECHNOLOGY FOR SEED AND FERTILISER PARTICLES AND ITS APPLICATION IN UAV-BASED SPREADING

引用

INMATEH - Agricultural engineering 2025年第1期75卷 584-599页

作者： Wang, Yanshuang Song, Cancan Gao, Yongqi Lan, Yubin Wang, Guobin Ming, Rui College of Agricultural Engineering and Food Science Shandong University of Technology Zibo China Shandong University of Technology Sub-center of National Center for International Collaboration Research on Precision Agricultural Aviation Pesticide Spraying Technology Zibo China Research of Institute of Ecological Unmanned Farm Shandong University of Technology Zibo China School of Computer and Big Data Minjiang University Fujian Provincial Key Laboratory of Information Processing and Intelligent Control Fuzhou China

The processes of sowing and fertiliser application represent a significant aspect of agricultural production. In order to achieve efficient and precise seeding and fertiliser application, mass flow detection of seed and fertiliser particles can facilitate real-time monitoring and precise decision-making for intelligent seeding and fertiliser application. However, the diversity of seed and fertiliser particle types and particle flow modes presents a challenge for existing detection methods, particularly in meeting the varying operational requirements, especially for particle mass flow detection in low-altitude and high-speed sowing operations of agricultural drones. In such cases, the overlap and flow rate will have a significant impact on the detection results due to the large displacement and the continuous high-throughput dense phase of the particle flow. This paper provides a summary of the existing seed and fertiliser particle mass flow detection techniques and their underlying working principles. It compares the direct detection based on mass with the indirect detection methods based on velocity and concentration, and analyses their respective advantages, disadvantages and applicability. It also explores the possibility of optimising the existing detection methods for the specific needs of agricultural UAVs and considers the potential introduction of cutting-edge science and technology in order to develop an efficient, accurate and convenient detection system to meet the growing market demand. © (2025), (INMA Bucharest). All Rights Reserved.

关键词： Precision agriculture

来源：评论

学校读者我要写书评

暂无评论

Improving hierarchical graph pooling with information bottleneck

引用

Journal of King Saud University Computer and Information sciences 2025年第3期37卷 1-15页

作者： Long, Jun Wang, Zidong Chen, Tingxuan Yang, Liu Big Data Institute Central South University Changsha China School of Computer Science and Engineering Central South University Changsha China

Hierarchical graph pooling has received significant attention in recent years due to its capability to capture hierarchical structures and deliver superior performance on graph-level tasks. However, most existing methods design pooling layers empirically and lack a unified theoretical framework. Consequently, the coarsened graphs generated by these methods frequently contain redundant information, leading to significant semantic bias. In this paper, we propose the principle of “minimal sufficient coarsened graph” to precisely define the desired properties of coarsened graphs. Guided by this principle, we introduce a novel Graph Pooling Information Bottleneck (GPIB) for hierarchical graph pooling. GPIB aims to achieve the desired attributes of minimality and sufficiency in coarsened graphs. Specifically, we first incorporate node degree properties to enrich node representations and introduce a dynamic generation mechanism. This mechanism regulates the information flow, ensuring substantial auxiliary knowledge for downstream tasks and enhancing information sufficiency. Then, we introduce a parameter-learning noise distribution, forcing the elimination of redundant information and thus, facilitating minimality and robustness in our model. Remarkably, we theoretically demonstrate that imposing constraints only on the coarsened graph produced by the final pooling layer equates to constraining all coarsened graphs. This insight simplifies the constraint procedure while ensuring its effectiveness. Extensive experiments on six real-world datasets validate the effectiveness of our proposed model, which consistently outperforms the baseline methods and produces high-quality coarsened graphs.

关键词： Graph classification Graph neural networks Graph pooling Graph representation learning

来源：评论

学校读者我要写书评

暂无评论

Recognition of American sign language using modified deep residual CNN with modified canny edge segmentation

引用

Multimedia Tools and Applications 2025年 1-28页

作者： Hariharan, U. Devarajan, Devikanniga Kumar, P. Santhosh Rajkumar, K. Meena, M. Akilan, T. Department of Computer Science and Engineering Apex Institute of Technology Chandigarh University Punjab Mohali140413 India Department of Computer Science and Engineering GITAM University Karnataka Bengaluru561203 India Department of Information Technology Faculty of Engineering and technology SRM Institute of Science and Technology Ramapuram Tamil Nadu Chennai600089 India Department of Data Science and Business Systems School of Computing Faculty of Engineering and Technology SRM Institute of Science and Technology Kattankulathur Tamil Nadu Chennai603203 India Department of Computer Science and Engineering Sri Sairam Engineering College Tamil Nadu Tambaram602109 India Research Coordinator Dept. of IQAC Amity University Kolkata West Bengal 700135 India

American Sign Language (ASL) recognition aims to recognize hand gestures, and it is a crucial solution to communicating between the deaf community and hearing people. However, existing sign language recognition algorithms still have some drawbacks, such as difficulty recognizing hand movements and low recognition accuracy for most sign language recognition. A Modified Convolutional Neural Network (MCNN) deep residual 101 classifier-based American Sign Language identification system has been developed to address this problem. There are three main parts presents in the method. The first part is pre-processing the images to remove the noise, enhance the contrast of the picture, adjust the contrast level and smoothen the picture using various filters. The second part is the segmentation, and it’s used to partition the image using a modified canny edge detection method by removing all weak edges present in the image. Finally, classification will be done using the Modified CNN deep residual 101 classifiers. By classifying the image, the American Sign Language is accurately identified. This process is conducted through images. The outcome shows that the suggested approach has a 0.97% accuracy and a 0.05% False Positive Rate. Other CNN architectures such as resNet 50 and resNet 18 reached 0.90% and 0.80% accuracy, respectively, which is lower than our proposed method. In addition, the resNet 101 classifier effectively recognizes the difficult hand gestures through the image data. It obtains high recognition accuracy for 36 signs is 0 to 9 numbers and a to z alphabets from American Sign Language. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2025.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Design, Performance, and Applications of AMMIS: A Novel Airborne Multimodular Imaging Spectrometer for High-Resolution Earth Observations

引用

engineering 2025年第4期47卷 38-56页

作者： Jia, Jianxin Wang, Yueming Zheng, Xiaorou Yuan, Liyin Li, Chunlai Cen, Yi Si, Fuqi Lv, Gang Wang, Chongru Wang, Shengwei Zhang, Changxing Zhang, Dong He, Daogang Zhuang, Xiaoqiong Han, Guicheng Zhang, Mingyang Hyyppä, Juha Wang, Jianyu Key Laboratory of Space Active Opto-Electronics Technology Shanghai Institute of Technical Physics Chinese Academy of Sciences Shanghai200083 China Department of Photogrammetry and Remote Sensing Finnish Geospatial Research Institute EspooFI-02150 Finland Guangdong Provincial Key Laboratory of Multimodal Big Data Intelligent Analysis Guangdong School of Computer Science and Engineering South China University of Technology Guangzhou510641 China State Key Laboratory of Remote Sensing Science Aerospace Information Research Institute Chinese Academy of Sciences Beijing100094 China Key Laboratory of Environmental Optics and Technology Anhui Institute of Optics and Fine Mechanics Hefei Institutes of Physical Science Chinese Academy of Sciences Hefei230031 China Department of Mechanical Engineering Aalto University EspooFI-02150 Finland

Airborne hyperspectral imaging spectrometers have been used for Earth observation over the past four decades. Despite the high sensitivity of push-broom hyperspectral imagers, they experience limited swath and wavelength coverage. In this study, we report the development of a push-broom airborne multimodular imaging spectrometer (AMMIS) that spans ultraviolet (UV), visible near-infrared (VNIR), shortwave infrared (SWIR), and thermal infrared (TIR) wavelengths. As an integral part of China's High-Resolution Earth Observation Program, AMMIS is intended for civilian applications and for validating key technologies for future spaceborne hyperspectral payloads. It has been mounted on aircraft platforms such as Y-5, Y-12, and XZ-60. Since 2016, AMMIS has been used to perform more than 30 flight campaigns and gather more than 200 TB of hyperspectral data. This study describes the system design, calibration techniques, performance tests, flight campaigns, and applications of the AMMIS. The system integrates UV, VNIR, SWIR, and TIR modules, which can be operated in combination or individually based on the application requirements. Each module includes three spectrometers, utilizing field-of-view (FOV) stitching technology to achieve a 40° FOV, thereby enhancing operational efficiency. We designed advanced optical systems for all modules, particularly for the TIR module, and employed cryogenic optical technology to maintain optical system stability at 100 K. Both laboratory and in-flight calibrations were conducted to improve preprocessing accuracy and produce high-quality hyperspectral data. The AMMIS features more than 1400 spectral bands, with spectral sampling intervals of 0.1 nm for UV, 2.4 nm for VNIR, 3 nm for SWIR, and 32 nm for TIR. In addition, the instantaneous fields of view (IFoVs) for the four modules were 0.5, 0.25, 0.5, and 1 mrad, respectively, with the VNIR module achieving an IFoV of 0.125 mrad in the high-spatial-resolution mode. This study reports on land

关键词： Fighter aircraft

来源：评论

学校读者我要写书评

暂无评论

A Reinforcement Learning Agent Controlled Multi-branch Small Object Detection Framework

A Reinforcement Learning Agent Controlled Multi-branch Small...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Hong, Junkun Long, Yitian Luo, Yueyi Hua, Liujie Long, Jun Qi, Qianqian School of Computer Science and Engineering Central South University Changsha China Data Science Institute Vanderbilt University NashvilleTN United States School of Mathematics and Statistics Central South University Changsha China Big Data Institution Central South University Changsha China

ISBN: (纸本)9798350368741

The past few years have witnessed the immense development of small object detection, which is aimed at detecting size-limited targets in high-resolution images. The prevailing methods focus on extracting fine-grained information by expanding the receptive fields and then generating the potential small object region. However, these solutions inevitably add sophisticated detectors and extra learning components, which incur time-consuming and computation-costing. Meanwhile, we observe that it's suboptimal to extract fine features in such a generic way. To alleviate the issues, we propose a multi-branch small object detection framework with a regular-scale detection branch and a small-scale detection branch. Specifically, we design and pre-train a reinforcement learning agent to control feature extractors in both branches according to the results of small object areas. Moreover, we present a region clipping algorithm to rebuild the small object to regular size, which can be input into a mature detector directly. The extensive experiments on COCO, VisDrone, SODA-D, and our collecting TVDS datasets demonstrate our method outperforms the state-of-the-art methods in several metrics. © 2025 IEEE.

关键词： multi-branch small object detection region clipping region discriminator reinforcement learning agent

来源：评论

学校读者我要写书评

暂无评论

Integration of Process Optimization and Automation: A Way to AI Powered Digital Transformation 1

Integration of Process Optimization and Automation: A Way to...

引用

1st International Conference on Advances in Computer science, Electrical, Electronics, and Communication Technologies, CE2CT 2025

作者： Kotte, Kulasekhara Reddy Thammareddi, Latha Kodi, Divya Anumolu, Venkateswara Rao Arun Kumar, K. Joshi, Shubham INC SLRI Solutions India India Cyber Security Senior Data Analyst Truist Financial India Technical Architect ERP Solutions- Information Technology India Big Data Architect India Bipin Tripathi Kumaon Institute of Technology Department of Computer Science & Engineering Almora India

ISBN: (纸本)9798331518578

One of the most important factors facilitating AI-powered digital transformation is the combination of process automation and optimisation. Organisations may improve operational efficacy, save costs, and streamline procedures by utilising cutting-edge technology like robotic process automation (RPA), machine learning (ML), and artificial intelligence (AI). While automation seeks to reduce human participation in repetitive and rule-based operations, process optimisation concentrates on improving current business processes to attain improved productivity. The development of intelligent, adaptable, and self-improving systems is made possible by the combination of these two methodologies. Predictive findings, real-time decision-making, and customised customer experiences are further made possible by AI-driven analytics. This connection promotes innovation, scalability, and agility across sectors, giving businesses a competitive edge in the quickly changing digital market. The article examines the main principles, advantages, and difficulties of this shift and provides a road map for businesses looking to automate and optimise processes using AI in order to attain operational excellence and sustainable growth. © 2025 IEEE.

关键词： Decision making

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：