检索结果-内蒙古大学图书馆

4th International Conference on Advancement in Electronics and Communication Engineering, AECE 2024

作者： Krishnamoorthy, A. Sundhar, K.A. Naveen Kumar, V. Karthik, V. Vellore Institute of Technology School of Computer Science & Engineering Tamil Nadu India Vellore Institute of Technology Department of Computer Science Engineering Tamil Nadu India

ISBN: (数字)9798350364729

ISBN: (纸本)9798350364729

Social media platforms have become essential avenues for individuals to express opinions and emotions across diverse topics such as products, events, and policies, using both text and emojis. Understanding these sentiments is crucial, yet manually analyzing the vast data on platforms like Twitter is impractical. Automated sentiment analysis tools, exemplified by Roberta, offer a solution. Roberta, an evolving language model, employs advanced bidirectional learning to predict sentiment labels accurately. This study focuses on sentiment analysis on Twitter, leveraging Roberta to autonomously categorize tweets into positive, negative, or neutral sentiments. We utilizing natural language processing and machine learning algorithms for dataset preprocessing, incorporating sources like Kaggle and ensuring data integrity. Assessment metrics encompassing accuracy, precision, recall, and F1-score, in addition to confusion metrics, assess model performance. According to our findings, machine learning powered sentiment analysis models are dependable, even in the face of challenges and more so when processing Twitter data sets. The study gives direction on how this kind of model can be of help in analyzing sentiments expressed through social networking sites thus assisting scholars and professional Consultancy in the field. © 2024 IEEE.

关键词： Tweets

来源：评论

学校读者我要写书评

暂无评论

Road Marking Detection in Challenging Environments using Optimized YOLOv7 3

Road Marking Detection in Challenging Environments using Opt...

引用

3rd International Conference on Machine Learning and data Engineering, ICMLDE 2024

作者： Gupta, Rajeev Kumar Pateriy, R.K. Swain, Debabrata Kumar, Abhijit Department of Computer Science and Engineering Pandit Deendayal Energy University Gujarat Gandhinagar India Department of Computer Science and Engineering Maulana Azad National Institute of Technology Bhopal India School of Computer Science UPES Dehradun India

Road marking detection area avails research with the application of computer vision and machine learning in the identification and analysis of various types of road markings such as lane markers, crosswalks, and road signs. It is very important to enhance safety on the roads by making these markings visible to drivers, cyclists, and pedestrians. Therefore, this study was designed to further enhance the efficiency and performance of the road marking detection models by incorporating the advanced AdamW optimizer and proposing a novel architecture known as Multi-Stage Cross-Convolutional Bottleneck Network-MSCCBN further-efficient in describing complex patterns and shapes from road markings. The testing was performed on a newly developed benchmark dataset containing 2,887 high-resolution images divided into training and test sets, with a total of 11 distinct road marking categories. Applying the Ceymo dataset and YOLOv7 detection framework, the result of the proposed model was an mAP50 score of 0.889, which means very high accuracy in object detection. This model indeed had a high precision of 0.886 and recall of 0.83, hence effectively minimizing both false positives and false negatives. © 2025 Elsevier B.V.. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Multilevel chaotic encryption model with cyclic redundancy check for medical data

引用

Neural Computing and Applications 2025年第5期37卷 3581-3603页

作者： Sharma, Suvita Rani Singh, Birmohan Kaur, Manpreet Department of Computer Science and Engineering Sant Longowal Institute of Engineering and Technology Punjab Longowal148106 India Department of Electrical and Instrumentation Engineering Sant Longowal Institute of Engineering and Technology Punjab Longowal148106 India

In telemedicine applications, it is crucial to ensure the authentication, confidentiality, and privacy of medical data due to its sensitive nature and the importance of the patient information it contains. Communication through open networks is insecure and has many vulnerabilities, making it susceptible to unauthorized access and misuse. Encryption models are used to secure medical data from unauthorized access. In this work, we propose a bit-level encryption model having three phases: preprocessing, confusion, and diffusion. This model is designed for different types of medical data including patient information, clinical data, medical signals, and images of different modalities. Also, the proposed model is effectively implemented for grayscale and color images with varying aspect ratios. Preprocessing has been applied based on the type of medical data. A random permutation has been used to scramble the data values to remove the correlation, and multilevel chaotic maps are fused with the cyclic redundancy check method. A circular shift is used in the diffusion phase to increase randomness and security, providing protection against potential attacks. The CRC method is further used at the receiver side for error detection. The performance efficiency of the proposed encryption model is proved in terms of histogram analysis, information entropy, correlation analysis, signal-to-noise ratio, peak signal-to-noise ratio, number of pixels changing rate, and unified average changing intensity. The proposed bit-level encryption model therefore achieves information entropy values ranging from 7.9669 to 8.000, which is close to the desired value of 8. Correlation coefficient values of the encrypted data approach to zero or are negative, indicating minimal correlation in encrypted data. Resistance against differential attacks is demonstrated by NPCR and UACI values exceeding 0.9960 and 0.3340, respectively. The key space of the proposed model is 1096, which is substantially mor

关键词： Telemedicine

来源：评论

学校读者我要写书评

暂无评论

An Iterative PRISMA Review of GAN Models for Image Processing, Medical Diagnosis, and Network Security

引用

computers, Materials & Continua 2025年第2期82卷 1757-1810页

作者： Uddagiri Sirisha Chanumolu Kiran Kumar Sujatha Canavoy Narahari Parvathaneni Naga Srinivasu Department of Computer Science and Engineering Prasad V Potluri Siddhartha Institute of TechnologyVijayawada520007India Department of Computer Science and Engineering Koneru Lakshmaiah Education FoundationVaddeswaram522302India Department of Electronics and Communication Engineering Sreenidhi Institute of Science and TechnologyHyderabad501301India Amrita School of Computing Amrita Vishwa VidyapeethamAmaravati522503India Department of Teleinformatics Engineering Federal University of CearáFortaleza60455-970Brazil Inti International University Nilai71800Malaysia

The growing spectrum of Generative Adversarial Network (GAN) applications in medical imaging, cyber security, data augmentation, and the field of remote sensing tasks necessitate a sharp spike in the criticality of review of Generative Adversarial Networks. Earlier reviews that targeted reviewing certain architecture of the GAN or emphasizing a specific application-oriented area have done so in a narrow spirit and lacked the systematic comparative analysis of the models’ performance metrics. Numerous reviews do not apply standardized frameworks, showing gaps in the efficiency evaluation of GANs, training stability, and suitability for specific tasks. In this work, a systemic review of GAN models using the PRISMA framework is developed in detail to fill the gap by structurally evaluating GAN architectures. A wide variety of GAN models have been discussed in this review, starting from the basic Conditional GAN, Wasserstein GAN, and Deep Convolutional GAN, and have gone down to many specialized models, such as EVAGAN, FCGAN, and SIF-GAN, for different applications across various domains like fault diagnosis, network security, medical imaging, and image segmentation. The PRISMA methodology systematically filters relevant studies by inclusion and exclusion criteria to ensure transparency and replicability in the review process. Hence, all models are assessed relative to specific performance metrics such as accuracy, stability, and computational efficiency. There are multiple benefits to using the PRISMA approach in this setup. Not only does this help in finding optimal models suitable for various applications, but it also provides an explicit framework for comparing GAN performance. In addition to this, diverse types of GAN are included to ensure a comprehensive view of the state-of-the-art techniques. This work is essential not only in terms of its result but also because it guides the direction of future research by pinpointing which types of applications require some

关键词： GAN CGAN WGAN DCGAN image analysis

来源：评论

学校读者我要写书评

暂无评论

Adapting to Evasive Tactics through Resilient Adversarial Machine Learning for Malware Detection 18

Adapting to Evasive Tactics through Resilient Adversarial Ma...

引用

18th INDIAcom;11th International Conference on Computing for Sustainable Global Development, INDIACom 2024

作者： Krishna, G. Bala Sravan Kumar, G. Ramachandra, Mummadi Sampurnima Pattem, K. Rani, D. Sandhya Kakarla, Geeta Cvr College of Engineering Department of Computer Science Engg. Rangareddy Ibrahimpatnam India Cvr College of Engineering Department of Computer Science Engg.-Cyber Security Rangareddy Ibrahimpatnam India Sreenidhi Institute of Science and Technology Department of Computer Science Engg. Hyderabad India Sreenidhi Institute of Science and Technology Department of Computer Science Engg. Cyber Security Hyderabad India

ISBN: (纸本)9789380544519

This paper presents the Adaptive Resilience-based Convolutional Network (ARCNet), a sophisticated machine learning framework specifically designed to detect advanced, evasive malware. ARCNet combines convolutional and recurrent neural networks, making it highly adaptable to changing cyber threats. Its core components, the Adversarial Learning Module (ALM), Predictive analysis Engine (PAE), and Dynamic Adaptation System (DAS), significantly boost its detection power. Tests using a synthetic dataset show ARCNet's superiority over traditional models like the Support Vector Machine (SVM). It achieved 95.2% accuracy under normal conditions (compared to SVM's 89.4%) and maintained 92.5% accuracy even during adversarial attacks (against SVM's 80.3%). Notably, ARCNet's detection rates improved from 78.5% to 86.7% in five months after integrating the DAS. These results confirm ARCNet's efficiency in tackling complex malware challenges, contributing greatly to cybersecurity. The study underscores the importance of evolving and enhancing machine learning methods to keep pace with the rapidly changing landscape of cyber threats. © 2024 Bharati Vidyapeeth, New Delhi.

关键词： Block chain Blockchains Challenges of Block chain Consensus algorithm data privacy Digital signatures Distributed ledger Faces Organizations Smart contracts

来源：评论

学校读者我要写书评

暂无评论

Enhanced EEG decoding with weight freezing in parallel deep neuro-fuzzy networks (PDNFN) for multi-neurological disorder diagnosis

引用

Iran Journal of computer science 2025年 1-16页

作者： Jain, Shraddha Srivastava, Rajeev Department of Computer Science and Engineering Indian Institute of Technology BHU Uttar Pradesh Varanasi 221011 India

This study introduces an improved EEG decoding model. It uses parameter freezing within a Parallel Deep Neuro-Fuzzy Network (PDNFN) to improve the accuracy and efficiency of diagnosing neurological disorders. The method outperforms other evaluated approaches, achieving an accuracy of 91.84%, precision of 92.15% and recall of 91.42%. Statistical validation-through p-values < 0.05 and confidence intervals-confirms the robustness of the results. Unlike traditional deep learning models that update all weights during training, this method freezes less important weights based on a mask matrix and input threshold. This technique encourages model sparsity, alleviates overfeeting, and reduces computational load without compromising key discriminative features. Compared to dropout, parameter freezing offers a more structures from of regularization, improving generalization in EEG analysis. The PDNFN architecture combines fuzzy logic with deep learning to handle the uncertainty in EEG signals and enhance feature learning. Experiments conducted across multiple publicly available EEG datasets demonstrate notable improvements in both classification performance and computational efficiency. Comparative evaluation further highlight its superiority over traditional fully connected networks and existing regularization methods. Overall, it provides a practical and efficient solution for real-time clinical diagnosis using EEG data. © The Author(s), under exclusive licence to Springer Nature Switzerland AG 2025.

关键词： EEG decoding Electroencephalogram Fuzzy logic Neurological disorders Parallel deep euro-fuzzy networks (PDNFN) Sparse connections Weight freezing

来源：评论

学校读者我要写书评

暂无评论

A Global Geometric analysis of Maximal Coding Rate Reduction 41

A Global Geometric Analysis of Maximal Coding Rate Reduction

引用

41st International Conference on Machine Learning, ICML 2024

作者： Wang, Peng Liu, Huikang Pai, Druv Yu, Yaodong Zhu, Zhihui Qu, Qing Ma, Yi Department of Electrical Engineering and Computer Science University of Michigan Ann Arbor United States Antai College of Economics and Management Shanghai Jiao Tong University Shanghai China Department of Electrical Engineering and Computer Science University of California Berkeley United States Department of Computer Science and Engineering The Ohio State University Columbus United States Institute of Data Science University of Hong Kong Hong Kong

The maximal coding rate reduction (MCR2) objective for learning structured and compact deep representations is drawing increasing attention, especially after its recent usage in the derivation of fully explainable and highly effective deep network architectures. However, it lacks a complete theoretical justification: only the properties of its global optima are known, and its global landscape has not been studied. In this work, we give a complete characterization of the properties of all its local and global optima, as well as other types of critical points. Specifically, we show that each (local or global) maximizer of the MCR2 problem corresponds to a low-dimensional, discriminative, and diverse representation, and furthermore, each critical point of the objective is either a local maximizer or a strict saddle point. Such a favorable landscape makes MCR2 a natural choice of objective for learning diverse and discriminative representations via first-order optimization methods. To validate our theoretical findings, we conduct extensive experiments on both synthetic and real data sets. Copyright 2024 by the author(s)

关键词： Network architecture

来源：评论

学校读者我要写书评

暂无评论

HYBRID DEEP LEARNING RECOMMENDATION SYSTEM FOR ACCURATE MOVIE AND PRODUCT REVIEW PREDICTIONS

引用

Scalable Computing 2025年第4期26卷 1902-1918页

作者： Dhawan, Sanjeev Kulvinder, Singh Manoj, Yadav Faculty of Computer Science and Engineering Department of Computer Science and Engineering University Institute of Engineering and Technology (U.I.E.T) Kurukshetra University Haryana Kurukshetra India Department of Computer Science and Engineering University Institute of Engineering and Technology (U.I.E.T) Kuruk-shetra University Haryana Kurukshetra India

This paper investigates the efficacy of deep learning models for sentiment analysis using two publicly available datasets: IMDb’s movie review dataset and Amazon’s product review dataset. The main objective was to evaluate the performance of various model architectures, particularly Long Short-Term Memory (LSTM) networks with dropout techniques, in emotion categorization under different settings. Key performance metrics, including accuracy, precision, recall, and F1 score, were used to train and validate several deep learning models: LSTM Spatial Dropout 1D, Bidirectional GRU-LSTM, Hybrid LSTM+GRU, and Bidirectional LSTM. The LSTM Spatial Dropout 1D model achieved remarkable results, with 93.00% accuracy and F1 score on the Amazon dataset, and an impressive 96.20% accuracy and 97.78% F1 score on the IMDb dataset. The Bidirectional GRU-LSTM model also performed exceptionally, achieving 98.69% accuracy, 96.16% precision, 94.62% recall, and 93.49% F1 score, outperforming many existing hybrid models in recommendation systems. By integrating forward and backward context, the Bidirectional GRU-LSTM model effectively captures complex temporal relationships, offering more accurate recommendations than traditional models that analyze data separately. This study underscores the robustness of LSTM-based architectures in sentiment analysis and highlights the potential of combining sentiment analysis with collaborative filtering to enhance precision and specificity in e-commerce recommendation systems. © 2025 SCPE.

关键词： Bi-LSTM Bi-LSTM- GRU Collaborative filtering Content-Based Filtering Deep Learning Recommendation Systems

来源：评论

学校读者我要写书评

暂无评论

A Systematic Review of Deep Learning Approaches for Vessel Segmentation in Retinal Fundus Images 2

A Systematic Review of Deep Learning Approaches for Vessel S...

引用

2nd International Conference on Artificial Intelligence, Computational Electronics and Communication System, AICECS 2023

作者： Hegde, Govardhan Prabhu, Srikanth Gupta, Shourya Prabhu, Gautham Manuru Palorkar, Anshita Srujan, Metta Venkata Bhandary, Sulatha V Department of Computer Science and Engineering India Department of Data Science and Computer Applications India Department of Information and Communication Technology Manipal Institute of Technology Karnataka Manipal576104 India Department of Ophthalmology Kasturba Medical College Karnataka Manipal576104 India

Retinal vessel segmentation (RVS) is important to accurately differentiate retinal vasculature for diagnosing and monitoring various ocular and systemic diseases. The traditional methods for RVS have mostly involved supervised learning, although semi-supervised and unsupervised techniques are on the rise. This paper reviews the increase in complexity of developments in RVS primarily after 2020. The methods were chosen to cover both the gradual transition over time and a variety of unorthodox or combinatorial approaches. This includes convolutional neural networks, encoder-decoder models, generative models, and other multi-modal or hybrid techniques. CNN approaches discussed employ Zero Phase Component analysis, Global Contrast Normalization, and reinforcement learning. Encoder-decoder models include approaches such as the use of skip and residual connections, spatial attention, and atrous enhancement U-Net. Generative models propose short link connections, recurrent residual blocks, and multi-scale features to refine convolutional blocks. Hybrid methods involve the use of connectivity features, the MISOdata Algorithm, cross-domain adaptation, and multiple filters (such as morphological, match, and Gabor). All the frameworks are compared based on their performance on the benchmark dataset DRIVE to provide a comprehensive understanding of the current state of RVS. © Published under licence by IOP Publishing Ltd.

关键词： Diagnosis

来源：评论

学校读者我要写书评

暂无评论

Federated Causally Invariant Feature Learning 39

Federated Causally Invariant Feature Learning

引用

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Guo, Xianjie Yu, Kui Cui, Lizhen Yu, Han Li, Xiaoxiao School of Computer Science and Information Engineering Hefei University of Technology China Key Laboratory of Knowledge Engineering with Big Data of Ministry of Education China School of Software Shandong University China College of Computing and Data Science Nanyang Technological University Singapore Department of Electrical and Computer Engineering The University of British Columbia Canada Vector Institute Canada

ISBN: (纸本)157735897X

Federated feature selection (FFS) is a promising field for selecting informative features while preserving data privacy in federated learning (FL) settings. Existing FFS methods focus on capturing the correlations between features and labels. They struggle to achieve satisfactory performance in the face of data distribution heterogeneity among FL clients, and cannot address the out-of-distribution (OOD) problem that arises when a significant portion of clients do not actively participate in FL training. To address these limitations, we propose Federated Causally Invariant Feature Learning (FedCIFL), a novel approach for learning causally invariant features in a privacy-preserving manner. We design a sample reweighting strategy to eliminate spurious correlations introduced by selection bias and iteratively estimate the federated causal effect between each feature and the labels (with the remaining features initially treated as confounders). By iteratively refining the confounding feature set to identify the true confounders, FedCIFL mitigates the impact of limited local data on the accuracy of federated causal effect estimation. Theoretical analysis proves the correctness of FedCIFL under reasonable assumptions. Extensive experiments on synthetic and real-world datasets demonstrate the superiority of FedCIFL against eight state-of-the-art baselines, beating the best-performing approach by 3.19%, 9.07% and 2.65% in terms of average test Accuracy, RMSE and F1 score, respectively. It is a first-of-its-kind FFS approach capable of handling Non-IID and OOD data simultaneously. The source code is available at https://***/Xianjie-Guo/FedCIFL. Copyright © 2025, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词： Federated learning

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：