检索结果-内蒙古大学图书馆

A Spatiotemporal machine learning Framework for the Prediction of Metocean Conditions in the Gulf of Mexico: Application to Loop Current and Loop Current Eddy Forecasting

A Spatiotemporal Machine Learning Framework for the Predicti...

引用

Offshore Technology conference, OTC 2025

作者： Steele, E.C.C. Juniper, M.C.R. Pillai, A.C. Ashton, I.G.C. Chen, J. Jaramillo, S. Zarate, L. Met Office FitzRoy Road Exeter United Kingdom University of Exeter Penryn Campus Penryn United Kingdom University of Plymouth Drake Circus Plymouth United Kingdom Shell Global Solutions US Inc HoustonTX United States

ISBN: (纸本)9781959025610

We are entering an exciting new era of data-driven weather prediction, where forecast models trained on historical data (including observations and reanalyses) offer an alternative to directly solving the governing equations of fluid dynamics. By capitalizing on a vast amount of available information – and capturing their inherent patterns that are not represented explicitly – such machine learning-based techniques have the potential to increase forecast accuracy, augmenting traditional physics-based equivalents. Here, we adapt and apply a promising machine learning framework – originally proposed by the present authors for regional prediction of ocean waves – to the operational forecasting of the Loop Current and Loop Current Eddies (LC/LCEs) in the Gulf of Mexico (GoM). The approach consists of using an attention-based long short-term memory recurrent neural network to learn the temporal patterns from a network of available observations, that is then combined with a random forest based spatial nowcasting model, trained on highresolution regional reanalysis data, to develop a complete spatiotemporal prediction for the basin. Since machine learning approaches are typically physics-agnostic, an identical framework to that developed for the prediction of ocean waves can be used for the prediction of surface currents, with the only difference being the training datasets to which this is exposed. This is illustrated here using a period of three months of training data from October 2022 to December 2022, with the model driven by only three observation sites in the northern GoM. As such, it is unrealistic to expect performance for an unseen week in January 2023 to be equivalent to smaller/simpler domains with a more favorable quantity, quality and coverage/ distribution of input observations but, despite these severe constraints, the ability of the model to forecast a plausible structure of the LC/LCE system is nonetheless impressive. The architecture of the MaLCOM framew

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Application of machine learning in Enterprise Financial Risk Assessment: A Study About China's A-Share Listed Manufacturing Companies 5th

Application of Machine Learning in Enterprise Financial Risk...

引用

5th international conference on Neural Computing for Advanced Applications (NCAA)

作者： Yul, Kexin Yul, Zengyi Ma, Shuomin Xu, Pan Zhejiang Univ Technol Hangzhou 310023 Peoples R China SpringAI LLC Bellevue WA 98007 USA

ISBN: (纸本)9789819770038;9789819770045

Based on the financial data of domestic A-share listed firms from 2000 to 2022, this paper aims to explore the effectiveness of machine learning in identifying the financial risk of Chinese A-share listed manufacturing firms. To this end, a variety of machine learning models, including logistic regression, decision tree, random forest, XGBoost, SVM, and LSTM, are used to assess the financial risk of enterprises, and the key attributes of enterprise financial risk are extracted through the interpretability exploration and importance coefficient measurement. From this paper, the following conclusions are drawn: (i) XGBoost model performs the best on all attributes, showing its strong ability in dealing with complex financial datasets, and LSTM, which adds time-series factors, performs poorly, which is speculated that it may be related to the incompatibility of the characteristics of the financial data, the reliance on the time-series features, and the need for financial fine features. (ii) Audit opinion, Net Profit and ROA are key influencing factors.

关键词： Financial risk machine learning XGBoost LSTM

来源：评论

学校读者我要写书评

暂无评论

Structured Process on FL for Big data Analysis 6

Structured Process on FL for Big Data Analysis

引用

6th international conference on Mobile Computing and Sustainable Informatics, ICMCSI 2025

作者： Govindaram, Anitha Prasath, J.S. Jayasakthi, K. Rajkumar, N. Porkodi, G. Jose Anand, A. Thandalam Chennai602105 India Department of Computer Science and Engineering Sapthagiri NPS University Chikkasandra Karnataka Bengaluru560057 India Department of ECE KCG College of Technology Karapakkam Tamil Nadu Chennai600097 India Department of CSE Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology Thiruvalluvar Tamil Nadu Avadi600062 India Department of Computer Science and Business Systems Vel Tech Multi Tech Dr. Rangaranjan Dr.Sakunthala Engineering College Avadi Tamil Nadu Thiruvallur600062 India

ISBN: (纸本)9798331522667

The integration of the Internet of Things (IoT) into various aspects of daily life is accelerated by the proliferation of cognitive capabilities and AI-driven applications. Traditional automated systems rely on centralized data storage and dispensation, which may be feasible in specific application scenarios. However, the increasing flexibility of recent IoT networks and rising concerns over data confidentiality are challenging this approach. Federated learning (FL), a decentralized collaborative AI technique, allows machine learning to occur directly on external IoT devices without the need for data to be centrally aggregated or shared. This paves the way for a wide variety of intelligent IoT applications to be developed. This article presents a thorough analysis of the along with a detailed analysis of the convergence of these two fields. After that, a full analysis of how FL is currently being utilized in several essential applications related to the IoT is provided. These include smart health, mobility, Unmanned Aerial Vehicles (UAVs), infrastructure improvements, and smart industries. In total, 49 publications from the years 2015-2022 were reviewed here. Finally, discuss future study challenges with issues and exposed questions about data with IoT. © 2025 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Digital twin enhanced with machine learning Algorithms for Irrigation Management Using Sensor data 6

Digital twin enhanced with Machine Learning Algorithms for I...

引用

6th international conference on Industry 4.0 and Smart Manufacturing, ISM 2024

作者： Carlo Tancredi, Giovanni Paolo Preite, Luca Vignali, Giuseppe CIPACK Center Parco Area delle Scienze 181/A Parma43124 Italy Department of Engineering for Industrial Systems and Technologies Parco Area delle Scienze 181/A Parma43124 Italy

Efficient irrigation management is critical for maximizing agricultural output while minimizing water waste. This study investigates the effectiveness of machine learning (ML) for data-driven irrigation decision support systems (DSS) to address the scalability issue in artificial intelligence applications. Therefore, the scientific literature highlights how scalability is the main issue that needs to be investigated. In this framework, a living lab focused on an olive cultivation in a greenhouse has been developed to train and test a digital twin application enhanced with machine learning algorithms. Specifically, the algorithm classifies the irrigation status by assessing soil and environmental sensor data. The integration between the physical and the virtual counterparts has been implemented exploiting a Message Queuing Telemetry Transport (MQTT) communication protocol, where soil and environmental data have been treated as input topics. The model performance has been evaluated by calculating accuracy, precision, recall, F1-score, and confusion matrices, which highlight the effectiveness of the model in managing irrigation by assessing no-crop specific features with the aim of improving the scalability in greenhouse applications. © 2024 The Authors. Published by Elsevier B.V.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

An Efficient Federated Meta Unlearning Algorithm with Enhanced Privacy Protection 20th

An Efficient Federated Meta Unlearning Algorithm with Enhan...

引用

20th IFIP WG 10.3 international conference on Network and Parallel Computing, NPC 2024

作者： Wang, Yani Ying, Zuobin Pan, Zijie Zhu, Enmin Zhou, Wanlei City University of Macau Macau S.A.R. Taipa China University of Macau Macau S.A.R. Taipa China

ISBN: (纸本)9789819628636

In machine learning, the ability for models to selectively forget specific data is increasingly crucial. Traditional "machine unlearning" methods often involve extensive retraining, which is computationally expensive and impractical for real-time applications. This paper introduces Federated Meta Unlearning (FMU), a novel approach within the Federated learning (FL) framework that merges the adaptability of meta-learning with the precision of machine unlearning. FMU utilizes a modified Model-Agnostic Meta-learning algorithm, employing model decomposition and weighted aggregation techniques. This approach enables efficient data influence removal while minimizing impact on overall model performance. The key innovation of FMU is the decomposition of the federated model into adjustable components, allowing targeted modifications or removal based on unlearning requests, followed by weighted re-aggregation to maintain model integrity and flexibility. Extensive experiments across diverse datasets show that FMU reduces the computational burden of traditional unlearning while preserving high model performance and adhering to privacy standards. Additionally, FMU enhances adaptability to new data and tasks, making it suitable for dynamic and privacy-sensitive settings. In summary, FMU offers a substantial advancement in addressing the challenge of selective forgetting in machine learning. © IFIP international Federation for Information Processing 2025.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

data Security Protection System in Human-machine-Object Fusion Intelligent Demand engineering System 3

Data Security Protection System in Human-Machine-Object Fusi...

引用

3rd international conference on Integrated Circuits and Communication Systems, ICICACS 2025

作者： Qi, Yongjun Li, Chenggao Tang, Hailin Shen, Junxuan Guangdong Baiyun University Faculty of Megadata and Computing Guangdong Guangzhou China School of Information and Communication Technology Mongolian University of Science and Technology Ulaanbaatar Mongolia

ISBN: (纸本)9798331508456

In the engineering system of human-machine physical intelligence requirements, the multi-source and high complexity of data pose significant security risks. Therefore, this article designs a comprehensive data security protection system. Firstly, this article chooses a dynamic encryption strategy based on role-based access control. Then, this article combines differential privacy algorithm to construct a privacy protection model to ensure the privacy of data during sharing and processing. Subsequently, this article proposes a trusted data transmission protocol. Finally, this article introduces federated learning techniques to further improve the security and robustness of data in distributed scenarios. The experimental results show that the system performs well in terms of data security, privacy protection, and processing efficiency. Especially under high-intensity attacks, the system successfully detected and prevented approximately 98% of tampering. In the above data conclusions, the research method has significant advantages in encryption performance, achieving a better balance between privacy protection and data accuracy, demonstrating strong application prospects. © 2025 IEEE.

关键词： Requirements engineering

来源：评论

学校读者我要写书评

暂无评论

Autonomous international Classification of Diseases Coding Using Pretrained Language Models and Advanced Prompt learning Techniques: Evaluation of an Automated Analysis System Using Medical Text

引用

JMIR MEDICAL INFORMATICS 2025年 13卷 e63020页

作者： Zhuang, Yan Zhang, Junyan Li, Xiuxing Liu, Chao Yu, Yue Dong, Wei He, Kunlun Chinese Peoples Liberat Army Gen Hosp Med Big Data Res Ctr 28 Fuxing Rd Beijing 100853 Peoples R China Beijing Inst Technol Sch Comp Sci & Technol Beijing Peoples R China Digital Hlth China Technol Co LTD Beijing Peoples R China Peoples Liberat Army Gen Hosp Med Ctr 6 Sr Dept Cardiol Beijing Peoples R China

Background: machine learning models can reduce the burden on doctors by converting medical records into international Classification of Diseases (ICD) codes in real time, thereby enhancing the efficiency of diagnosis and treatment. However, it faces challenges such as small datasets, diverse writing styles, unstructured records, and the need for semimanual preprocessing. Existing approaches, such as naive Bayes, Word2Vec, and convolutional neural networks, have limitations in handling missing values and understanding the context of medical texts, leading to a high error rate. We developed a fully automated pipeline based on the Key-bidirectional encoder representations from transformers (BERT) approach and large-scale medical records for continued pretraining, which effectively converts long free text into standard ICD codes. By adjusting parameter settings, such as mixed templates and soft verbalizers, the model can adapt flexibly to different requirements, enabling task-specific prompt learning. Objective: This study aims to propose a prompt learning real-time framework based on pretrained language models that can automatically label long free-text data with ICD-10 codes for cardiovascular diseases without the need for semiautomatic preprocessing. Methods: We integrated 4 components into our framework: a medical pretrained BERT, a keyword filtration BERT in a functional order, a fine-tuning phase, and task-specific prompt learning utilizing mixed templates and soft verbalizers. This framework was validated on a multicenter medical dataset for the automated ICD coding of 13 common cardiovascular diseases (584,969 records). Its performance was compared against robustly optimized BERT pretraining approach, extreme language network, and various BERT-based fine-tuning pipelines. Additionally, we evaluated the framework's performance under different prompt learning and fine-tuning settings. Furthermore, few-shot learning experiments were conducted to assess the feasibil

关键词： BERT bidirectional encoder representations from transformers pretrained language models prompt learning ICD international Classification of Diseases cardiovascular disease few-shot learning multicenter medical data

来源：评论

学校读者我要写书评

暂无评论

Research on Fracture Propagation of Fractured Shale Reservoirs by Supercritical CO2 Fracturing Based on Numerical Simulation and machine learning Tools 30th

Research on Fracture Propagation of Fractured Shale Reservoi...

引用

30th international conference on Computational and Experimental engineering and Sciences

作者： Han, Lei Shi, Xian Ni, Hongjian Zhang, Weidong Ge, Xiaoxin China Univ Petr East China Coll Petr Engn Qingdao 266580 Peoples R China China Univ Petr East China State Key Lab Deep Oil & Gas Qingdao 266580 Peoples R China

ISBN: (纸本)9783031774911;9783031774898;9783031774881

Understanding the interaction between fracturing fractures and natural fractures (NF) is crucial for estimating fracture complexity and simulated reservoir volume (SRV). For the simulation of complex network models, the general time cost is enormous, therefore, it is very meaningful to consider using machine learning algorithms to predict fracture propagation. On the basis of establishing a fluid structure coupling model using the Cohesive module of ABAQUS and completing indoor experimental verification, this article introduces the global embedded cohesion zone model (CZM) combined with the natural fracture (NF) model (generated in Python) to establish a dispersed mesh model. Based on this model, compare the fracture propagation patterns of hydraulic fracturing and supercritical CO2 fracturing. Design experimental schemes to study the problem of fracture propagation under different engineering and geological conditions. Using machine learning (LSTM-Long Short-Term Memory algorithm) to predict the changes in fracture length under different time and conditions, and comparing it with actual simulation results to optimize the algorithm. The results indicate that supercritical CO2 is more inclined to penetrate natural fractures, leading to the expansion of fractures within the rock matrix. Hydraulic fracturing fractures combine well with natural fractures. The established prediction model can effectively predict the state of fracture propagation. This study combines numerical simulation and machine learning, which is of great significance for enriching the theory of supercritical CO2 fracturing.

关键词： Shale reservoir Natural fracture Supercritical CO2 Cohesion zone model LSTM model

来源：评论

学校读者我要写书评

暂无评论

Performance Evaluation of machine learning Models for Predicting Voltage Swell Peak Amplitude in Grid-tied Photovoltaic Systems 33

Performance Evaluation of Machine Learning Models for Predic...

引用

33rd Southern African Universities Power engineering conference, SAUPEC 2025

作者： May, Nontlahla Muremi, Lutendo Bokoro, Pitshou May, Siyasanga Innocent Hlaluku Mkasi, Wisani University of Johannesburg Department of Electrical and Electronic Engineering Technology Johannesburg South Africa Energy Centre Council for Scientific and Industrial Research Pretoria South Africa

ISBN: (纸本)9798331535162

This research investigates the effectiveness of machine learning and deep learning models in forecasting voltage swell peak amplitudes within grid-connected photovoltaic (PV) systems, aiming to enhance power quality management. A 24-month dataset (January 2022 - December 2023) encompassing power and weather data from a 3.3 kWp PV system at the Council for Scientific and Industrial Research (CSIR) in Pretoria, South Africa, was utilized. Hourly averaged data between 5 am and 6 pm, capturing PV system and weather measurements, was analysed. Five models - Artificial Neural Network (ANN), Support Vector Regression (SVR), Random Forest (RF), K-Nearest Neighbours (KNN), and Long Short-Term Memory (LSTM) - were trained and evaluated using Mean Squared Error (MSE) and Root Mean Squared Error (RMSE). The Random Forest model demonstrated superior predictive accuracy, closely aligning with actual peak voltages and achieving the lowest MSE (0.01V2) and RMSE (0.02V). This study highlights the potential of machine learning, particularly Random Forest, in accurately predicting voltage swell peak amplitudes, contributing to improved power quality management in grid-connected PV systems. © 2025 IEEE.

关键词： Mean square error

来源：评论

学校读者我要写书评

暂无评论

Study on State-of-the-Art machine learning Algorithms Efficiency for CO2 Emissions Prediction in Fleet Management Optimization 18th

Study on State-of-the-Art Machine Learning Algorithms Effici...

引用

18th international conference on Interdisciplinarity in engineering, INTER-ENG 2024

作者： Muraru, Mădălina Maria Simó, Zsuzsa Iantovics, László Barna Pharmacy Science and Technology of Targu Mures George Emil Palade University of Medicine Targu Mures540142 Romania

ISBN: (纸本)9783031816840

Global carbon dioxide emissions from fossil fuels and industry reached a record in 2022, emphasizing the importance of reducing pollution from the automotive sector. Vehicles represent a significant source of CO₂ emissions, and large fleets are a major challenge in managing the carbon footprint. Although there are strict regulations and initiatives to reduce emissions, companies that manage large fleets face difficulties in collecting and reporting accurate data on vehicle emissions. The lack of uniformity in the provision of data by car manufacturers and the difficulties associated with manual retrieval of information from vehicle documents complicate this process. To address these challenges and with the aim of selecting the best algorithm to integrate into a decision support system, the following machine learning algorithms were involved in the research: Random Forest, XGBoost, and a Feed-Forward Neural Network, to predict CO₂ emissions based on vehicle characteristics from a fleet of over 26,500 records. The models were trained using numerical and categorical data, properly preprocessed, providing a good solution for emissions prediction in the absence of real measured data for newly ordered vehicles. Based on these findings, all of them can be integrated into a complex decision support system, enabling more effective and informed decision-making. The results indicate that the Random Forest algorithm is the most effective for predicting CO2 emissions in this context, as it captures complex relationships between features and the target variable delivering accurate predictions. In this context, the use of machine learning models for large vehicle fleets not only ensures CO2 emission predictions but also contributes to sustainable fleet management and compliance with environmental regulations. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Low emission

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：