检索结果-内蒙古大学图书馆

15th international conference on Soft Computing and Pattern Recognition, SoCPaR 2023 and 14th World Congress on Nature and Biologically Inspired Computing, NaBIC 2023

作者： Tank, Dharmesh R. Patel, Sanjay G. Pandya, Devang S. CE Department LDRP-ITR Kadi Sarva Vishwavidyalaya Gujarat Gandhinagar 382016 India CSE Department Nirma University Gujarat Ahmedabad 382481 India IT Department Ganpat University Gujarat Mehsana 384012 India

ISBN: (纸本)9783031810824

Autonomous crowd activity analysis is imperative in light of recent global events. The study of crowd behavior has generated considerable interest in the fields of computer vision and cognitive science. Consider a scenario where a security guard is tasked with monitoring a multitude of CCTV camera feeds, particularly those used in police investigations. To effectively manage, count, secure, and track a crowd occupying a shared space, such as during COVID-19 outbreaks and public events. Assessing crowd situations presents challenges due to substantial occlusion, intricate actions, and variations in posture. The development of an algorithm capable of monitoring each video and automatically identifying suspicious activity would enhance the accuracy and efficiency of a large-scale intelligent system. This serves as the driving force behind video anomaly detection. The fundamental concept involves training a model on typical activities using training videos featuring normal behavior, and subsequently utilizing those activities that are different from any normalized seen in the training video. The comprehensive inclusion of all conceivable future anomalous events within the training video is unattainable due to the inherent limitations associated with the complete knowledge and capture of such occurrences. Deep learning, a branch of machine learning, employs artificial neural networks to acquire knowledge from data. Notably, deep learning has been successfully employed in the field of crowd anomaly detection, which poses significant challenges but holds great significance. Deep learning has demonstrated considerable potential for addressing this problem and has surpassed conventional machine learning systems in terms of performance. This review aims to present an exhaustive analysis of video anomaly detection systems based on deep learning techniques that were introduced since year 2019. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Anomaly Detection Crowd Behaviour Analysis dataset Deep learning Approaches Surveillance Video

来源：评论

学校读者我要写书评

暂无评论

Enhanced Diabetes Prediction Through Advanced machine learning and Imbalance Handling Techniques

Enhanced Diabetes Prediction Through Advanced Machine Learni...

引用

Computing and Information Technology (ICCIT-1441), 2020 international conference on

作者： Inam Abousaber Dept. of Information Technology Faculty of Computers and Information Technology University of Tabuk Tabuk Saudi Arabia

ISBN: (数字)9798350353839

ISBN: (纸本)9798350353846

This paper presents a comprehensive analysis of various machine learning classifiers for diabetes prediction, with a specific focus on addressing the challenges posed by imbalanced datasets. The study evaluates the effectiveness of three popular imbalance handling techniques, Adaptive Synthetic (ADASYN), Synthetic Minority Over-sampling Technique (SMOTE), and Random Under-Sampling, when applied to different classifiers, including Gradient Boosting, Decision Tree, K-Nearest Neighbors, and Logistic Regression. Through extensive experimentation on the PIMA Indians Diabetes dataset and the Diabetes dataset 2019, the proposed methodology integrates advanced feature engineering and robust imbalance handling to enhance prediction accuracy. The results demonstrate that the Gradient Boosting classifier, particularly when combined with the ADASYN technique, consistently outperforms other models across all evaluation metrics, achieving state-of-the-art performance. This study not only highlights the importance of selecting appropriate imbalance handling methods but also provides practical insights into the strengths and limitations of different classifiers in the context of medical data classification. The findings are further supported by visualizations such as ROC curves, precision-recall curves, and confusion matrices, offering a detailed understanding of model behavior. This research contributes to the development of more accurate and reliable predictive models for diabetes, with potential applications in broader medical data analysis.

关键词： Analytical models Accuracy data analysis Computational modeling Predictive models Boosting Reliability engineering data models Diabetes Context modeling

来源：评论

学校读者我要写书评

暂无评论

Comparative Study Using Ensemble Methods and Sampling Techniques for Imbalanced Diabetes data 4th

Comparative Study Using Ensemble Methods and Sampling Techni...

引用

4th international conference on Intelligent Systems and Pattern Recognition, ISPR 2024

作者： Khemisa, Hana Dendani, Nadjette Khedimi, Yasmine Amara, Nour Djihane Azizi, Nabiha Department of Computer Science Badji Mokhtar University Annaba Algeria

ISBN: (纸本)9783031821523

In medical decision making, the efficacy of classification, a vital component of decision support systems, can be significantly compromised by imbalanced class distributions within training data. This discrepancy is especially critical due to the heightened consequences of misclassification, particularly when a minority class is mistakenly classified as a majority class. Diabetes, posing a substantial public health challenge, underscores the urgency of addressing classification issues. The World Health Organization reports that 30% to 40% of diabetes cases remain undiagnosed, contributing to a 3% increase in diabetes-related mortality rates from 2000 to 2019. In 2019 alone, approximately 2 million deaths were attributed to diabetes and related kidney diseases. Consequently, effective intervention strategies are imperative. Managing imbalanced datasets in medical contexts presents significant obstacles to accurate classification, with profound implications for patient outcomes. To address this challenge, this research proposes leveraging machine learning and artificial intelligence techniques. The research comprises two main components: applying boosting and unbalanced bagging to a dataset containing diabetes-related information. Initial findings indicate that boosting yields promising results in classification accuracy. Building upon this success, we compare boosting with three additional technique Synthetic Minority Over-sampling Technique (SMOTE), Edited Nearest Neighbors (ENN), and Tomek links each sequentially applied to the dataset. Through rigorous experimentation and analysis, we aim to identify the most effective strategy for handling imbalanced datasets in diabetes classification. By enhancing our understanding of classification methods and their impact on imbalanced datasets, this research aims to contribute to improved decision-making processes in healthcare settings, ultimately leading to better outcomes for patients affected by diabetes. © The Author(s)

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Enhancing Network Security: An Ensemble of machine learning Model for Detection of Distributed Denial of Service Attack 2nd

Enhancing Network Security: An Ensemble of Machine Learning ...

引用

2nd international conference on Emerging Wireless Technologies and sciences, ICEWTS 2024

作者： Rajput, Deepak Singh Upadhyay, Arvind Kumar Amity University Madhya Pradesh Gwalior India

ISBN: (纸本)9783031878855

This research aims to address the pressing issue of detecting and mitigating distributed denial of service (DDoS) attacks on computer networks. A hybrid machine learning model is proposed that leverages unsupervised and supervised algorithms to classify network traffic as normal or malicious with high accuracy. The CICDDoS2019 dataset containing real and simulated attack traffic is used to train and evaluate various classification algorithms. Feature selection is applied to extract the most relevant attributes for detection. The proposed ensemble model combines K-Means clustering, random forest, extreme gradient boosting, adaptive boosting, support vector machine, and artificial neural network classifiers. Experimental results demonstrate the model achieves 99.65% overall accuracy in distinguishing DDoS attacks, outperforming individual algorithms. This research contributes towards developing more robust techniques for analyzing network behaviors and enhancing intrusion detection capabilities. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Creating and Analyzing a Multimedia dataset for Building Energy Efficiency Estimation 4th

Creating and Analyzing a Multimedia Dataset for Building Ene...

引用

4th international conference on Smart Multimedia, ICSM 2024

作者： Chen, Gong Qi Kashyap, Ninad Zhang, Zihao Kucheva, Yana Bobker, Michael Zhu, Zhigang Department of Computer Science The CUNY Graduate Center 365 5th Avenue New YorkNY10016 United States Sustainability in the Urban Environment The CUNY City College 160 Convent Avenue New YorkNY10031 United States Spitzer School of Architecture The CUNY City College 160 Convent Avenue New YorkNY10031 United States Department of Sociology The CUNY City College 160 Convent Avenue New YorkNY10031 United States Building Performance Laboratory The CUNY City College 160 Convent Avenue New YorkNY10031 United States Department of Computer Science The CUNY City College 160 Convent Avenue New YorkNY10031 United States

ISBN: (纸本)9783031824746

This paper presents the results of a research that created and analyzed a Multimedia dataset for building energy efficiency estimation. First a new Multimedia Building Energy Efficiency (MMBEE) dataset was created from publicly available data. This work then explored the use of the window-to-wall ratio (WWR) information from building facade images and integrated it with traditional tabular data to create new training data, in order to predict building energy efficiency measures. Finally, we discuss potential applications and future research directions in using the MMBEE dataset for building energy efficiency prediction. Throughout the paper, a number of important processes and analyses were performed, which include feature selection, data correlation analysis, WWR extraction, and comparison of deep network and random forest models in building energy efficiency estimation. From this first attempt at using the Multimedia dataset for building energy efficiency estimation, we found the performances of deep models were better than traditional models such as random forest. We also found that there was an optimal point of what features shall be used for the prediction. Nonetheless, the incorporation of the current WWR estimation results did not yield the anticipated enhancement in estimation performance. Subsequently, a comprehensive investigation was conducted to ascertain potential contributing factors, and several avenues for future research were identified to enhance the predictive utility of the WWR feature. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Building energy efficiency computer vision machine learning Multimedia databases window-to-wall ratio

来源：评论

学校读者我要写书评

暂无评论

The Evaluation of Adversarial Attacks Against ML-Powered NIDS in a Realistic Scenario 2nd

The Evaluation of Adversarial Attacks Against ML-Powered NID...

引用

2nd international conference on applied Intelligence, ICAI 2024

作者： Uccello, Federica Pawlicki, Marek Pawlicka, Aleksandra D’Antonio, Salvatore Kozik, Rafal Choraś, Michal University of Naples ‘Parthenope’ Centro Direzionale Isola C4 Naples80133 Italy Linköping University Linköping Sweden Bydgoszcz University of Science and Technology PBS Bydgoszcz Poland ITTI Poznań Poland

ISBN: (纸本)9789819619061

machine learning-based Network Intrusion Detection Systems have significantly advanced network security by identifying and mitigating sophisticated threats. However, current research on adversarial attacks focuses on pre-aggregated datasets rather than raw traffic, which impacts the realism of the scenario. This paper evaluates a more realistic adversarial attack scenario by manipulating raw traffic in the PCAP files before the aggregation step of NetFlow data. This method better simulates real-world conditions, where traffic is captured and processed dynamically. We implement and compare traditional adversarial attacks using the Fast Gradient Sign Method directly on NetFlow data with the realistic scenario of infecting raw PCAP traffic. The main goal is to demonstrate whether such adversarial attacks would maintain the same effectiveness when performed on real traffic. Experimental results demonstrate that the realistic approach maintains the integrity and functionality of network traffic, and manages to affect the NIDS performance. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Intrusion Detection in the Internet of Vehicles Using Transformer Models

Intrusion Detection in the Internet of Vehicles Using Transf...

引用

Computational Intelligence Approaches and Applications (ICCIAA), international conference on

作者： Mohammad Alauthman Ashraf Mashaleh Nauman Aslam Amjad Aldweesh Ammar Almomani Department of Information Security University of Petra Jordan Al-Balqa Applied University Al-Salt Jordan Department of Computer and Information Sciences Northumbria University U.K. College of computing and IT Shaqra University Shaqra Saudi Arabia Department of Computer Information Science Higher Colleges of Technology UAE Al-Balqa Applied University Jordan

ISBN: (数字)9798331523657

ISBN: (纸本)9798331523664

The proliferation of Internet of Vehicles (IoV) systems has introduced critical cybersecurity vulnerabilities in connected vehicle infrastructures. This paper presents a novel application of Transformer architecture for intrusion detection in vehicular Controller Area Network (CAN) buses, specifically addressing Denial-of-Service and Spoofing attacks. We introduce a Transformer-based model optimized for CAN frame sequence analysis, evaluated on the CICIoV2024 dataset comprising real-world attack scenarios from a 2019 Ford vehicle. Our experimental results demonstrate superior detection capabilities compared to classical machine learning approaches and recurrent neural architectures, achieving 98.4% accuracy and 98.5% F1-score. The proposed architecture's self-attention mechanism effectively captures temporal dependencies in CAN frame sequences while maintaining computational efficiency (2.3ms inference time) suitable for automotive-grade ECUs. This research advances the state-of-the-art in IoV security through enhanced detection accuracy and practical deployment considerations for resource-constrained vehicular environments.

关键词： Analytical models Adaptation models Computational modeling Intrusion detection Computer architecture machine learning Transformers Controller area networks Internet of Vehicles Automotive engineering

来源：评论

学校读者我要写书评

暂无评论

The application of ultrasonic measurement and machine learning technique to identify flow regime in a bubble column reactor

引用

Sensors international 2025年第1期6卷 100294-100294页

作者： Wongsaroj, Wongsakorn Thong-Un, Natee Hansot, Jirayut Shoji, Naruki Treenuson, Weerachon Kikura, Hiroshige Department of Instrumentation and Electronics Engineering King Mongkut's University of Technology North Bangkok 1518 Wongsawang Bangsue Bangkok10800 Thailand Center of Excellence on Instrumentation Technology and Automation King Mongkut's University North Bangkok Prachinburi Campus 129 Noenhom MuangPrachinburi25230 Thailand Robotics and Mechanical Engineering Research Unit Muroran Institute of Technology 27-1 Mizumoto-cho MuroranHokkaido050-0071 Japan Office of Atoms for Peace Ministry of Science and Technology 2-12-1 Ookayama Meguro-ku Tokyo152-8550 Japan Laboratory for Zero-Carbon Energy Tokyo Institute of Technology 2-12-1 Ookayama Meguro Tokyo152-8550 Japan

This paper presents a novel technique to classify the flow regimes in bubble columns. The ultrasonic velocity profiler is employed to detect the velocity deviation and echo characteristic of bubbles rising in the column. This information is set as attribute data for the machine learning algorithm. Classification-based machine learning is utilized to classify the flow regimes: bubbly, transition, and churn turbulent, which are defined as categories of the algorithm. Several classifiers were applied in this work, such as K-nearest neighbors, Decision tree, Support vector machines, Naive bayes, and Logical regression. The experimental demonstration was conducted to verify the performance of the proposed technique. Three kinds of two-phase flow with stagnant liquid that had various viscosities were used for the experiment. The air within the superficial velocity range was injected to alter the flow regime. The flow regime classification model was set. The proposed method was applicable to identify the flow regimes. The classifiers were tested, and their accuracy was evaluated. © 2024 The Authors

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Deep learning accelerated inverse modeling and forecasting for large-scale geologic CO2 sequestration

引用

international Journal of Greenhouse Gas Control 2025年 144卷

作者： Chen, Bailian Yan, Bicheng Aslam, Billal Kang, Qinjun Harp, Dylan Pawar, Rajesh Earth and Environmental Sciences Division Los Alamos National Laboratory NM87545 United States Physical Science and Engineering Division King Abdullah University of Science and Technology Thuwal23955 Saudi Arabia The Freshwater Trust OR97205 United States

Traditional physics-simulation based approaches for inverse modeling and forecasting in geologic CO2 sequestration (GCS) are very time consuming. For example, a single inverse modeling may take a few weeks for a large-scale CO2 storage model without leveraging any high-performance computing. To speed up this process, we developed a novel approach that employs machine learning methods to integrate monitoring data into subsurface forecasts more rapidly than current physics-based inverse modeling workflows allow. These updated forecasts with the updated models from the inverse modeling process will be used to provide site operators with decision support by generating real-time performance metrics of CO2 storage (e.g., CO2 plume and pressure area of review). First, we developed a deep learning (DL) model to predict the pressure/saturation evolution in large-scale storage reservoirs. A feature coarsening technique was applied to extract the most representative information and perform the training and prediction at the coarse scale, and to further recover the resolution at the fine scale by 2D piecewise cubic interpolation. The accuracy of the feature coarsening-based DL model is validated with a reservoir model built upon a Clastic Shelf storage site. Thereafter, the feature coarsening-based DL model was utilized as forward model in the inverse modeling process where a classical data assimilation approach, ES-MDA-GEO, was applied. The efficiency and effectiveness of the proposed DL-assisted workflow for large-scale inverse modeling and forecasting was demonstrated with the Clastic Shelf storage model. © 2025 Elsevier Ltd

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

Advancements on multi-fidelity random Fourier neural networks: application to hurricane modeling for offshore wind energy

Advancements on multi-fidelity random Fourier neural network...

引用

AIAA science and Technology Forum and Exposition, AIAA SciTech Forum 2025

作者： Davis, Owen Geraci, Gianluca Wentz, Jacqueline M. King, Ryan N. Cortiella, Alexandre Rybchuk, Alex Gomez, Miguel Sanchez Deskos, Georgios Motamed, Mohammad Optimization and Uncertainty Quantification Sandia National Laboratories AlbuquerqueNM United States Computational Data Science Sandia National Laboratories LivermoreCA United States National Wind Technology Center National Renewable Energy Laboratory GoldenCO United States Department of Applied Mathematics and Statistics University of New Mexico AlbuquerqueNM United States

ISBN: (数字)9781624107238

ISBN: (纸本)9781624107238

Multi-fidelity approaches are emerging as effective strategies in computational science to handle otherwise intractable tasks like Uncertainty Quantification (UQ), training of machine learning (ML) models, and optimization, for expensive high-fidelity applications in which the amount of available simulations or data is limited. The main idea is simple: large datasets generated for low-fidelity approximations of the problem at hand are fused with a much sparser dataset for the target (high-fidelity) system. In this paper, we build on our recent success in designing random Fourier Neural Networks (rFNNs) [1] to target problems arising in wind energy applications and in particular problems of interest for hurricane modeling. In this context, data for the high-fidelity models are limited and lower fidelity alternatives are needed. In this work, we introduce a novel multi-fidelity training approach for our rFNNs and demonstrate its use on a simple verification problem and on a hurricane modeling problem in which high-fidelity data are generated via Large-Eddy Simulations (LES), while low-fidelity data are given by a mesoscale model. Initial results demonstrate how the multi-fidelity training approach can improve the quality of the resulting surrogate. © 2025, American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved.

关键词： Large eddy simulation

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：