检索结果-内蒙古大学图书馆

4th International Conference on Emerging Systems and Intelligent Computing, ESIC 2024

作者： Das, Surajit Nayak, Samaleswari P. Sahoo, Biswajit Nayak, Sarat Chandra Kiit Deemed to Be University School of Computer Engineering Bhubaneswar India Silicon Institute of Technology Dept. of Computer Science and Engineering Bhubaneswar India Gitam Deemed to Be University School of Technology Department of Computer Science and Engineering Hyderabad India

ISBN: (纸本)9798350349856

The complexities of using ensemble models to rectify class imbalance in healthcare datasets are investigated in this study. Using well-known methods such as Random Forest, Gradient Boosting, AdaBoost, CatBoost, LightGBM, XGBoost, BRF and EE we perform an extensive analysis on data-sets that exhibit different levels of class imbalance. The study uses stratified k-fold cross-validation for robust evaluation and the Synthetic Minority Over-sampling Technique (SMOTE) for data augmentation. ROC curves are shown, and performance indicators like as accuracy, precision, recall, F1-score, and area under the curve (AUC) are examined. The results shed light on the complex effects of ensemble models on unbalanced healthcare data-sets and offer directions for further investigation. © 2024 IEEE.

关键词： Adaptive boosting

来源：评论

学校读者我要写书评

暂无评论

Improving the review classification of Google apps using combined feature embedding and deep convolutional neural network model

引用

Journal of Ambient Intelligence and Humanized Computing 2023年第4期14卷 4257-4272页

作者： Aslam, Naila Alzamzami, Ohoud Xia, Kewen Sadiq, Saima Umer, Muhammad Bisogni, Carmen Ashraf, Imran School of Electronics and Information Engineering Hebei University of Technology Tianjin China Faculty of Computing and Information Technology King Abdulaziz University Jeddah Saudi Arabia Department of Computer Science Khwaja Fareed University of Engineering and Information Technology Rahim Yar Khan Pakistan Department of Computer Science Information Technology The Islamia University of Bahawalpur Bahawalpur Pakistan Department of Computer Science University of Salerno Fisciano Italy Information and Communication Engineering Yeungnam University Gyeongsan Korea Republic of

Online reviews play an integral part in making mobile applications stand out from the large number of applications available on the Google Play store. Predominantly, users consider posted reviews for appropriate app selection. Manual categorization of such reviews is both inefficient and time-consuming. Therefore, automatic analysis of the sentiments of such reviews provides fast suggestions for new users and facilitates their selection of the appropriate app. However, data imbalance is a major challenge for performing class prediction of such reviews as their distribution is sparse and often leads to low accuracy. This work proposes a framework to overcome this limitation. Extensive experiments are performed using the original and balanced data with the synthetic minority oversampling technique (SMOTE) and adaptive synthetic sampling (ADASYN). Additionally, deep learning and machine learning models are evaluated using FastText, FastText Subword, global vector (GloVe), and their combinations for word representation. Baseline machine learning models, including random forest, extra tree classifier, gradient boosting, Naive Bayes, logistic regression (LR), stochastic gradient descent (SGD), and voting classifier (VC) that combines LR and SGD, are used for comparison. The outcomes show that the convolutional neural network using a combination of word embedding techniques produces the most accurate results. © 2023, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Real-Time Facial Recognition System for Secure College Bus Transport using Deep Learning Techniques 3

Real-Time Facial Recognition System for Secure College Bus T...

引用

3rd International Conference on Electronics and Renewable Systems, ICEARS 2025

作者： Kavitha, M.N. Saranya, S.S. Pragatheeswari, E. Priya, N. Kaviya Ranjith, S.R. Vanchivel, N. Rahul Kongu Engineering College Department of Computer Science and Design Tamil Nadu India Psg Institute of Technology and Applied Research Department of Computer Science and Engineering Tamil Nadu Coimbatore India Kongu Engineering College Department of Computer Science and Engineering Tamil Nadu Erode India Kongu Engineering College Department of Csd Tamil Nadu India

ISBN: (纸本)9798331509675

The proposed system aims to enhance student transportation security through real-time face detection and recognition. Leveraging the MTCNN framework for accurate face detection and the FaceNet model for reliable face recognition, the system ensures only authorized students can board designated buses by comparing captured faces with a pre-registered database. In case of mismatches, alerts are triggered to notify drivers and authorities. Additionally, parallel implementation of Support Vector Machine (SVM) and K-Nearest Neighbors (KNN) classifiers improves authentication accuracy. The system's deep learning-based architecture ensures robust performance under varying environmental conditions and supports continuous learning for long-term reliability. © 2025 IEEE.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

Advanced Motion Recognition Using High-Resolution Infrared Imaging and Enhanced Image Processing Techniques 3

Advanced Motion Recognition Using High-Resolution Infrared I...

引用

3rd International Conference on Disruptive Technologies, ICDT 2025

作者： Mishra, Prafull Srivastava, Rishabh Singh, Prerna Shrivastava, Vineet Garg, Devesh Raj Kumar Goel Institute of Technology Department of Computer Science & Engineering UP Ghaziabad India

ISBN: (纸本)9798331519582

Over the past few years, infrared (IR) motion detection has become more important in domains such as security, wildlife monitoring, and nighttime surveillance, when standard vision-based systems frequently fail. The approach of infrared video detection using convolutional neural networks is been implemented in this research work. Infrared imaging is particularly useful for collecting temperature-based data in low-visibility or dark conditions, enabling for the monitoring of activities that would otherwise be difficult to trace. The suggested approach preprocesses grayscale infrared pictures, normalizing them and scaling to fit the input requirements of a custom-built CNN. Through a sequence of convolutional and pooling layers, the CNN collects characteristics related to motion patterns in IR pictures, learning to discriminate between frames having motion and those that do not. The proposed system was developed with the help of pre-trained data values of infrared pictures, and performance was tested using multiple parameters that includes various visualization figures such as accuracy-precision graph, recall, and ROC-AUC, to show its usefulness. This real-time approach combines a CNN with a video processing pipeline to identify motion in an infrared movie. It uses visualizations like as confusion matrices, accuracy-loss curves, and precision-recall curves to give insights into model design, implementation details, and experimental results. The results reveal that CNNs can accurately discern motion in infrared pictures, paving the way for automated low-light surveillance. © 2025 IEEE.

关键词： Thermography (imaging)

来源：评论

学校读者我要写书评

暂无评论

A link-quality-aware grouping-based efficient routing scheme in Software-Defined Vehicular Networks 17

A link-quality-aware grouping-based efficient routing scheme...

引用

17th International Conference on COMmunication Systems and NETworkS, COMSNETS 2025

作者： Malakar, Madhuri Mahapatro, Judhistir National Institute of Technology Rourkela Department of Computer Science and Engineering Odisha Rourkela769008 India

ISBN: (纸本)9798331531195

This paper delves into the existing routing schemes in Vehicular Ad hoc Networks (VANETs), critically examines their limitations, and proposes an efficient routing scheme for Software-Defined Vehicular Networks (SDVN). The proposed scheme employs a grouping mechanism to establish a robust path between the source and destination vehicles by using velocity vectors and cartesian coordinates of vehicles. Additionally, the scheme incorporates real-time vehicle information, such as GPS location, direction, speed, and vehicle density, and the Expected Transmission Count (ETX), as a link-quality routing metric for superior link quality. The ETX metric ensures the selection of better and more stable links, while the grouping mechanism identifies links with a high probability of breakage prior to their occurrence for efficient routing. The real-time link-quality information between the vehicles is dynamically generated and monitored at the control plane, ensuring up-to-date and accurate routing decisions. We evaluate the performance of the proposed scheme against traditional approaches. Simulation results show that the proposed scheme outperforms the existing schemes, achieving higher throughput, lower packet loss, reduced end-to-end delay, and significantly fewer path breaks with the network scenario of increasing vehicle density and speed. © 2025 IEEE.

关键词： Vehicular ad hoc networks

来源：评论

学校读者我要写书评

暂无评论

NeuroSpectrum Deep Learning Architecture for Automated Spectrum Analysis Resource Optimization 3

NeuroSpectrum Deep Learning Architecture for Automated Spect...

引用

3rd International Conference on Integrated Circuits and Communication Systems, ICICACS 2025

作者： Tamilselvan, K. Senthil, T. Buvaneswari, B. Santhiya, S. Soni, Mukesh Avs Engineering College Department of Information Technology Salem India Sreenivasa Institute of Technology and Management Studies Department of Computer Science and Engineering chittoor India Panimalar Engineering College Department of Information Technology Chennai India Excel Engineering College Department of Computer Science and Engineering Namakkal India Lovely Professional University Department of Division of Research and Development Phagwara India

ISBN: (纸本)9798331508456

Reducing the availability of spectrum, as well as the nature of allocating channels in the present wireless telecommunication networks, is one of the critical issues facing modern wireless technology. Such spectrum sensing methods may not be highly flexible or efficient in environments where the radio spectrum undergoes frequent changes. This work proposes an improved deep learning spectrum sensing model based on CNN with reinforcement learning. The proposed system design calls for a pioneering dual mechanism, incorporating both spectrum analysis and a resource allocation learning model. The framework achieved remarkable performance metrics: 96 The spectrum sensing method achieved a 3% occurrence rate, a 94.8% precision rate for resource allocation, and a 92.7% efficiency in dynamic spectrum use. In the present investigation, we have employed it for different topologies and got 45% improvement in spectrum efficiency and 87% less interference than the conventional techniques. The work challenges deep learning-based spectrum sensing as well as management and indicates that they are prepared for the following generation of the wireless communication network. © 2025 IEEE.

关键词： Cognitive radio

来源：评论

学校读者我要写书评

暂无评论

Sample-based relationship for assisting diagnosis of pneumonia in medical care

引用

Multimedia Tools and Applications 2025年第1期84卷 267-288页

作者： Chen, Hongkang Lu, Huijuan Zhu, Wenjie Zhou, Ye Yao, Yudong Wang, Renfeng College of Information Engineering China Jiliang University Zhejiang Hangzhou China Drore Information and Technology Co. LTD Zhejiang Hangzhou China Department of Electrical and Computer Engineering Stevens Institute of Technology HobokenNJ United States College of Computer Science and Technology Zhejiang University Zhejiang Hangzhou China

Deep neural networks have succeeded in learning balanced and imbalanced data in the field of pneumonia diagnosis. However, both require separate model designs in their respective domains. The pneumonia recognition methods have poor transfer performance between the two types of image distribution data. This paper proposes BeFormer, which focuses on sample relationships. BeFormer introduces two stages, feature extraction and establishing sample relationships, to solve the problem. Specifically, in the feature extraction stage, a new backbone network called PANet based on Multi-Scale Frequency Domain Attention (FPSA) is proposed. In establishing sample relationships, Deformable Transformer is used in parallel to model the relationships between partial regions of interest among the same and different samples in the pneumonia dataset. These two modules are connected in series to form BeFormer. Without making any changes to the model, this paper conducted experiments using BeFormer on both balanced and imbalanced pneumonia datasets. As compared to traditional pneumonia recognition models, the proposed method achieved significant improvements in the two different data distributions, demonstrating outstanding transfer performance. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Advance Path Loss Model for Distance Estimation Using LoRaWAN Network's Received Signal Strength Indicator (RSSI)

引用

IEEE Access 2024年 12卷 83205-83216页

作者： Vo, Hoang Hoang Long Nguyen, Van Tran, Van Lic Ferrero, Fabien Lee, Fang-Yi Tsai, Meng-Hsun The University of Danang-University of Science and Technology Faculty of Advance Science and Technology Da Nang550000 Viet Nam The University of Danang-University of Science and Technology Department of Electronic and Telecommunication Engineering Da Nang550000 Viet Nam LEAT CNRS Universite Cote d'Azur Sophia Antipolis06903 France National Cheng Kung University Department of Computer Science and Information Engineering Tainan70101 Taiwan National Yang Ming Chiao Tung University Department of Computer Science Hsinchu30010 Taiwan

This work introduces a novel approach to improve the precision of distance estimation in localization systems by using existing LoRaWAN and RSSI-based techniques. Despite the benefits of range and power efficiency, these systems exhibit limited accuracy in practical situations. To address the limitation, this study provides an innovative technique that greatly improves the precision of distance estimations, particularly in urban environments. The fundamental basis of this approach lies in the use of a dynamic path loss model. An additional element is to accommodate the varied and dynamic conditions of signal transmission in metropolitan areas. A better Kalman filter is also used in the study. This is important because it reduces the effects of multipath fading and environmental noise that often make RSSI-based localization in LoRaWAN networks less accurate. The study further examines the influence of the environmental exponent, also known as the path loss exponent, on the RSSI results and the precision of the distance measurements. This methodology achieves the average error under 1 meters for indoor environments and under 7 meters for outdoor environments. Finally, the Cumulative Density Function (CDF) shows 90 % of the distance estimation algorithm error for indoor environment is lower than 1.08 meters while for outdoor environment is lower than 7.55 meters. Based on these improvements, the introduced methodology not only enhances and improves existing approaches but also optimizes the precision and dependability of urban localization technologies, with substantial implications for a variety of practical applications. © 2013 IEEE.

关键词： Kalman filters

来源：评论

学校读者我要写书评

暂无评论

Anticipating the Nearness of Coronary Heart Infection Utilizing Machine Learning Classifiers 2

Anticipating the Nearness of Coronary Heart Infection Utiliz...

引用

2nd International Conference on Machine Learning and Data engineering, ICMLDE 2023

作者： Salem Akoosh, Lamiaa Mohammed Siddiqui, Farheen Zafar, Sherin Naaz, Sameena Alam, M Afshar Department of Computer Science and Engineering School of Engineering Sciences and Technology Jamia Hamdard New Delhi India

Researchers are putting in a lot of time and energy into developing methods to use machine learning algorithms, a subfield of AI, to diagnose disease in an individual patient. Extensive studies have been conducted on the potential benefits of using machine learning techniques in the treatment of cardiovascular disease. For this preliminary study, the study have zeroed down on heart illness to get as specific as possible about our methodology. In this paper, the authors investigate the differences in accuracy between different machine learning approaches specifically, (SVM), (KNN), and a(ANN) when applied to the categorization of cardiovascular illness. Our research makes use of approximately seventy thousand patient records from the Kaggle dataset, which focuses on cardiovascular disease The Kaggle dataset comprises a limited number of variables per patient record, encompassing serum cholesterol, diastolic and systolic blood pressure, relative blood glucose levels, and the presence or absence of angina. The present study examines the Kaggle dataset and employs the (KNN), (SVM), and (ANN) methodologies. The results indicate that the K-nearest neighbors (KNN) algorithm, specifically with a value of 9 for the number of neighbors, achieved an accuracy of ?0.997. The SVM model using default hyperparameters achieved an accuracy of 0.9997. The Support Vector Machine (SVM) model utilizing a Radial Basis Function (RBF) kernel and a C value of 100.0 exhibited an accuracy of 0.9998. Conversely, the SVM model employing a linear kernel and a C value of 1000.0 achieved a perfect accuracy of 1.000. The Feedforward neural network achieved a perfect accuracy of 1.000 using the Adam optimization algorithm after 10 epochs and 50 *** future research, the authors intend to employ a deep neural network hypermodel in order to enhance the accuracy of their findings. © 2024 Elsevier B.V.. All rights reserved.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Security challenges of IoT with its applications and architecture

Security challenges of IoT with its applications and archite...

引用

International Conference on Artificial Intelligence, Blockchain, Computing and Security, ICABCS 2023

作者： Bathla, Neha Kaur, Amanpreet Department of Computer Science and Engineering Chitkara University Institute of Engineering & Technology Chitkara University Punjab India

ISBN: (纸本)9781032669663

IoT (Internet of Things) is the network of physical objects("things") in which embedded devices are used like sensors and software that are used for connecting the devices and objects over the internet. Currently, seven billion connected IoT devices today, by 2025 experts is expecting this number to grow to twenty-two billion. In this article we provide an introduction or overview of the IoT (Internet of Things), IoT architecture and functionality of each layer, its applications and security issues. The IoT are enabled by the newest developments in smart sensors and RFID. The basic principle of IoT is that it has smart sensors that work together directly without human participation to deliver the new applications. In the upcoming years, with the intelligent decision making IoT expected as bridge diverse technologies to enable new applications by connecting physical objects together. In this survey we also describe the importance of security and its challenges. © 2024, CRC Press/Balkema. All rights reserved.

关键词： Internet of things

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：