检索结果-内蒙古大学图书馆

(r, k, Ε)-Anonymization: Privacy-Preserving Data Publishing Algorithm Based on Multi-Dimensional Outlier Detection, k-Anonymity, and Ε-Differential Privacy

引用

IEEE Access 2025年 13卷 70422-70435页

作者： Kara, Burak Cem Eyupoglu, Can Karakus, Oktay National Defence University Turkish Air Force Academy Department of Computer Engineering İstanbul34149 Turkey Cardiff University School of Computer Science and Informatics CardiffCF24 4AG United Kingdom

In recent years, there has been a tremendous rise in both the volume and variety of big data, providing enormous potential benefits to businesses that seek to utilize consumer experiences for research or commercial purposes. The general data protection regulation (GDPR) implementation, on the other hand, has introduced extensive control over the use of individuals’ personal information and placed many limits. Data anonymization technologies have become an important solution for businesses trying to generate value from data while adhering to GDPR limitations. To address these challenges, researchers have developed various methods, including k-anonymity and Ε -differential privacy, offering solutions for both industry and academia. However, protecting individuals’ privacy against diverse attack attempts presents significant challenges for anonymization models that rely solely on a single technique, highlighting the need for more adaptable and hybrid approaches. In this study, a new hybrid anonymization algorithm called (r, k, Ε )-anonymization has been proposed, which combines k-anonymity and Ε-differential privacy models in a consistent framework and provides stronger privacy guarantees compared to existing privacy-preserving models. The proposed algorithm is capable of overcoming well-known shortcomings of the k-anonymity and Ε -differential privacy models, and it has been confirmed by extensive tests on real-world datasets. The proposed (r, k, Ε )-anonymization algorithm outperforms k-anonymity and Ε -differential privacy in terms of the average error rate measure, achieving data utility increases of 31.74% and 26.99%, respectively. © 2013 IEEE.

关键词： Anonymity

来源：评论

学校读者我要写书评

暂无评论

Addressing Imbalance in Health Datasets: A New Method NR-Clustering SMOTE and Distance Metric Modification

引用

computers, Materials & Continua 2025年第2期82卷 2931-2949页

作者： Hairani Hairani Triyanna Widiyaningtyas Didik Dwi Prasetya Afrig Aminuddin Department of Electrical Engineering and Informatics Faculty of EngineeringUniversitas Negeri MalangMalang65145Indonesia Department of Computer Science Universitas BumigoraMataram83127Indonesia Department of Computer Graphic and Multimedia Faculty of ComputingCollege of Computing and Applied SciencesUniversiti Malaysia Pahang Al-Sultan AbdullahPekan26600Malaysia

An imbalanced dataset often challenges machine learning, particularly classification methods. Underrepresented minority classes can result in biased and inaccurate models. The Synthetic Minority Over-Sampling Technique (SMOTE) was developed to address the problem of imbalanced data. Over time, several weaknesses of the SMOTE method have been identified in generating synthetic minority class data, such as overlapping, noise, and small disjuncts. However, these studies generally focus on only one of SMOTE’s weaknesses: noise or overlapping. Therefore, this study addresses both issues simultaneously by tackling noise and overlapping in SMOTE-generated data. This study proposes a combined approach of filtering, clustering, and distance modification to reduce noise and overlapping produced by SMOTE. Filtering removes minority class data (noise) located in majority class regions, with the k-nn method applied for filtering. The use of Noise Reduction (NR), which removes data that is considered noise before applying SMOTE, has a positive impact in overcoming data imbalance. Clustering establishes decision boundaries by partitioning data into clusters, allowing SMOTE with modified distance metrics to generate minority class data within each cluster. This SMOTE clustering and distance modification approach aims to minimize overlap in synthetic minority data that could introduce noise. The proposed method is called “NR-Clustering SMOTE,” which has several stages in balancing data: (1) filtering by removing minority classes close to majority classes (data noise) using the k-nn method;(2) clustering data using K-means aims to establish decision boundaries by partitioning data into several clusters;(3) applying SMOTE oversampling with Manhattan distance within each cluster. Test results indicate that the proposed NR-Clustering SMOTE method achieves the best performance across all evaluation metrics for classification methods such as Random Forest, SVM, and Naїve Bayes, compared t

关键词： SMOTE modification Clustering-SMOTE manhattan distance

来源：评论

学校读者我要写书评

暂无评论

Deep Learning Quantitative Susceptibility Mapping With Reduced Training Time Based on Local Field Transformation

引用

IEEE Transactions on Emerging Topics in Computational Intelligence 2025年第2期9卷 1527-1538页

作者： Wu, Ming-Long Huang, De-Rong Cheng, Jhih-Shan National Cheng Kung University Department of Computer Science and Information Engineering Tainan701 Taiwan National Cheng Kung University Institute of Medical Informatics Tainan701 Taiwan

A quantitative susceptibility mapping (QSM) approach using single-orientation imaging data is proposed in this study. The proposed method generates local field maps at five predefined orientations via deep learning from single-orientation local field maps, followed by physical dipole inversion using the analytical Calculation of Susceptibility through Multiple Orientation Sampling (COSMOS) algorithm. We termed the method pseudo COSMOS (pCOSMOS). By isolating the physical dipole inversion process from deep learning, training of the neural network model becomes greatly accelerated. Brain imaging data from 10 healthy subjects (7 training and 3 testing) acquired at 3T with five different orientations relative to the main magnetic field were included to evaluate the reconstruction quality. Results compared against two iterative (Morphology Enabled Dipole Inversion or MEDI, and Structural Feature-based Collaborative Reconstruction or SFCR) and two deep learning QSM approaches (QSMnet and Learned Proximal Convolutional Neural Network or LPCNN) using COSMOS as the reference standard revealed that the susceptibility maps obtained from pCOSMOS yielded global quantitative performance comparable to QSMnet and LPCNN, while being prominently superior to MEDI and SFCR. Estimation errors in magnetic susceptibility were between 0.01 and 0.02 ppm voxel-wise, and 0.006 ± 0.004 ppm for five selected gray matter areas. Visual inspection further suggested better preserved tissue boundaries in pCOSMOS than QSMnet and LPCNN. Training time was reduced from 2 days for LPCNN and 12 hours for QSMnet to 3.2 hours for pCOSMOS. It is concluded that the proposed pCOSMOS method is an alternative deep learning QSM approach when computational efficiency is essential. © 2017 IEEE.

关键词： Brain mapping

来源：评论

学校读者我要写书评

暂无评论

Attention-Driven Emotion Recognition in EEG: A Transformer-Based Approach With Cross-Dataset Fine-Tuning

引用

IEEE Access 2025年 13卷 69369-69394页

作者： Ghous, Ghulam Najam, Shaheryar Alshehri, Mohammed Alshahrani, Abdulmonem AlQahtani, Yahya Jalal, Ahmad Liu, Hui Riphah International University Department of Electrical and Computer Engineering Islamabad44000 Pakistan Bahria University Department of Electrical Engineering Islamabad44000 Pakistan King Khalid University Department of Computer Science Abha61421 Saudi Arabia King Khalid University Department of Informatics and Computer Systems Abha61421 Saudi Arabia Air University Faculty of Computer Science and AI Islamabad44000 Pakistan Korea University College of Informatics Department of Computer Science and Engineering Seoul02841 Korea Republic of University of Bremen Cognitive Systems Laboratory Bremen28359 Germany Guodian Nanjing Automation Company Ltd. Nanjing210032 China

Emotion recognition from EEG (electroencephalogram) signals is crucial in mental health diagnostics and human-computer interaction but is often hindered by high dimensionality, noise, and complex temporal dependencies in the data. This paper presents a novel approach that integrates transformer models, attention mechanisms, and transfer learning to enhance emotion recognition accuracy from EEG signals. The proposed methodology consists of two phases: Attention Enhanced Base Model Development (AE-BMD) and Cross-Dataset Fine Tuning Adaptation (CD-FTA). In the AE-BMD phase, the base model is developed and trained on the SEED-IV dataset (15 participants, 62 EEG channels), achieving an accuracy of 84%, with an average precision of 84.75%, recall of 84% and F1-score of 84%. This phase employs optimized feature extraction from key EEG frequency bands (Delta, Theta, Alpha, Beta, Gamma) using techniques such as MFCC, GFCC, power spectral density, and Hjorth parameters. A transformer encoder with integrated spectral and temporal attention mechanisms captures intricate patterns and long-range dependencies within the EEG signals. In the CD-FTA phase, the model undergoes fine-tuning on the SEED-V dataset (20 participants, 62 channels) leading to an improved accuracy of 90%, with an average precision of 90.6%, recall of 90.6%, and F1-score of 90.6%. The model’s generalization is further validated on the MPED dataset (23 participants, 62 channels, seven emotion classes), achieving 79%, with an average precision of 79.3%, recall of 79.3% and F1-score of 79.1% across diverse emotional states. This cross-dataset adaptation leverages transfer learning to enhance the model’s generalization across different emotional states and EEG datasets. Experimental results show that the proposed approach outperforms traditional methods, achieving superior accuracy and robustness in emotion recognition tasks. This work advances emotion recognition systems by addressing challenges in EEG signal proc

关键词： Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

An AI-Enabled Framework for Transparency and Interpretability in Cardiovascular Disease Risk Prediction

引用

computers, Materials & Continua 2025年第3期82卷 5057-5078页

作者： Isha Kiran Shahzad Ali Sajawal ur Rehman Khan Musaed Alhussein Sheraz Aslam Khursheed Aurangzeb Department of Computer Science University of Engineering and TechnologyLahore39161Pakistan Dipartimento di Farmacia e Biotecnologie Alma Mater Studiorum—Universita di BolognaBologna40126Italy Department of Information Sciences University of EducationLahore54770Pakistan Department of Computer Science National Textile UniversityKarachi CampusKarachi74900Pakistan Department of Computer Science National Textile UniversityFaisalabad37610Pakistan Department of Computer Engineering College of Computer and Information SciencesKing Saud UniversityRiyadh11543Saudi Arabia Department of Electrical Engineering Computer Engineeringand InformaticsCyprus University of TechnologyLimassol3036Cyprus Department of Computer Science CTL EurocollegeLimassol3077Cyprus

Cardiovascular disease(CVD)remains a leading global health challenge due to its high mortality rate and the complexity of early diagnosis,driven by risk factors such as hypertension,high cholesterol,and irregular pulse *** diagnostic methods often struggle with the nuanced interplay of these risk factors,making early detection *** this research,we propose a novel artificial intelligence-enabled(AI-enabled)framework for CVD risk prediction that integrates machine learning(ML)with eXplainable AI(XAI)to provide both high-accuracy predictions and transparent,interpretable *** to existing studies that typically focus on either optimizing ML performance or using XAI separately for local or global explanations,our approach uniquely combines both local and global interpretability using Local Interpretable Model-Agnostic Explanations(LIME)and SHapley Additive exPlanations(SHAP).This dual integration enhances the interpretability of the model and facilitates clinicians to comprehensively understand not just what the model predicts but also why those predictions are made by identifying the contribution of different risk factors,which is crucial for transparent and informed decision-making in *** framework uses ML techniques such as K-nearest neighbors(KNN),gradient boosting,random forest,and decision tree,trained on a cardiovascular ***,the integration of LIME and SHAP provides patient-specific insights alongside global trends,ensuring that clinicians receive comprehensive and actionable *** experimental results achieve 98%accuracy with the Random Forest model,with precision,recall,and F1-scores of 97%,98%,and 98%,*** innovative combination of SHAP and LIME sets a new benchmark in CVD prediction by integrating advanced ML accuracy with robust interpretability,fills a critical gap in existing *** framework paves the way for more explainable and transparent decision-making in he

关键词： Artificial Intelligence cardiovascular disease(CVD) explainability eXplainable AI(XAI) interpretability LIME machine learning(ML) SHAP

来源：评论

学校读者我要写书评

暂无评论

RGB-D Video Mirror Detection

RGB-D Video Mirror Detection

引用

2025 IEEE/CVF Winter Conference on Applications of computer Vision, WACV 2025

作者： Xu, Mingchen Herbert, Peter Lai, Yu-Kun Ji, Ze Wu, Jing School of Computer Science and Informatics Cardiff University United Kingdom School of Engineering Cardiff University United Kingdom

ISBN: (纸本)9798331510831

Mirror detection aims to identify mirror areas in a scene, with recent methods either integrating depth information (RGB-D) or making use of temporal information (video). However, utilizing both data is still under-explored due to the lack of a high-quality dataset and an effective method for the RGB-D Video Mirror Detection (DVMD) problem. To the best of our knowledge, this is the first work to address the DVMD problem. To exploit depth and temporal information in mirror segmentation, we first construct a large-scale RGB-D Video Mirror Detection Dataset (DVMD-D), which contains 17977 RGB-D images from 273 diverse videos. We further develop a novel model, named DVMDNet, which can first locate the mirrors based on triple consistencies: local consistency, cross-modality consistency and global consistency, and then refine the mirror boundaries through content discontinuity, taking the temporal information within videos into account. We conduct a comparative study on the DVMD dataset, evaluating 12 state-of-the-art models (including single-image mirror detection, single-image glass detection, RGB-D mirror detection, video shadow detection, video glass detection, and video mirror detection methods). Code is available from https://***/UpChen/2025_DVMDNet. © 2025 IEEE.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

End-to-end optimized ensemble model for Alzheimer’s disease detection 19

End-to-end optimized ensemble model for Alzheimer’s disease...

引用

19th International Conference on Ubiquitous Information Management and Communication, IMCOM 2025

作者： Hassan, Abdelrahman Abuhmed, Tamer El-Sappagh, Shaker Faculty of Computer Science and Engineering Galala University Egypt College of Computing and Informatics Sungkyunkwan University Korea Republic of

ISBN: (纸本)9798331507817

Alzheimer's disease (AD) is a progressive neurodegenerative disorder with an increasing prevalence among the elderly, making early and accurate diagnosis critical for effective intervention and management. This paper introduces an end-to-end machine learning pipeline optimized for detecting AD using multimodal data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). The study proposes an ensemble model that integrates key modalities—demographics, cognitive scores, neuropsychological assessments, and MRI data—via early fusion to enhance AD detection. Various machine learning models, including Logistic Regression, Gradient Boosting, AdaBoost, XGBoost, Support Vector Machine, and Neural Networks, are compared, and feature selection is optimized through Recursive Feature Elimination, ANOVA F-value, and mutual information methods, with mutual information showing the most efficacy. With further hyperparameter tuning, the highest accuracy is achieved at 95.85%, with Gradient Boosting as the top-performing model. Additional experiments assess and evaluate individual modality contributions, underscoring the predictive value of cognitive scores and neuropsychological features. To enhance clinical interpretability, we incorporate model explainability using LIME and SHAP, which identify key features influencing model predictions across different AD stages. Explainable AI (XAI) is essential in healthcare for providing transparency and supporting clinician confidence by highlighting which biomarkers, such as MMSE, CDRSB, and PTAU, contribute most to diagnostic outcomes. This study demonstrates classical machine learning's potential in AD diagnostics, offering a pathway to cost-effective and clinically viable models for neurodegenerative disease management. © 2025 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Leveraging Large Language Models for Smart Pharmacy Systems: Enhancing Drug Safety and Operational Efficiency 19

Leveraging Large Language Models for Smart Pharmacy Systems:...

引用

19th International Conference on Ubiquitous Information Management and Communication, IMCOM 2025

作者： Osheba, Abdelrahman Abou-El-Ela, Ahmed Adel, Osama Maaod, Youssef Abuhmed, Tamer El-Sappagh, Shaker Faculty of Computer Science and Engineering Galala University Egypt College of Computing and Informatics Sungkyunkwan University Korea Republic of

ISBN: (纸本)9798331507817

Adverse drug reactions (ADRs) remain a crucial challenge in healthcare systems, highly contributing to patient mortality. We present an innovative smart pharmacy system that utilizes advanced large language models (LLMs) to enhance drug safety and pharmacy operational efficiency. Our system integrates real-time data from patient's prescriptions, medication databases, and electronic health records (EHR) to automate the detection of potential drug interactions, optimizing clinical decision-making and reducing ADR-related risks. The system's architecture is designed to easily manage prescription processing for patients, inventory management and control, and Pharmacist consultations through an intelligent chatbot interface. Key features include real-time tracking of medication expiration and inventory levels, an interaction checker API for identifying and mitigating risky drug combinations, and an LLM-powered chatbot for accurate data analysis and visualization. By combining advanced computational techniques with AI-driven insights, our smart pharmacy system not only improves medication safety but also eases pharmacy operations. This transformative approach holds the potential to significantly reduce ADR-related hospital admissions and enhance overall healthcare delivery. Our research underscores the vital role of AI and LLMs in modern pharmacy practice, offering an inclusive solution that integrates easily into existing healthcare infrastructures. The proposed smart pharmacy system can be plugged into hospital EHR systems to automatically track patient medications. © 2025 IEEE.

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

Design and In-training Optimization of Binary Search ADC for Flexible Classifiers 25

Design and In-training Optimization of Binary Search ADC for...

引用

30th Asia and South Pacific Design Automation Conference, ASP-DAC 2025

作者： Lozano Duarte, Paula Carolina Afentaki, Florentia Zervakis, Georgios Tahoori, Mehdi Department of Computer Science Karlsruhe Inst. of Tech. Karlsruhe Germany Computer Engineering and Informatics Department Univ. of Patras Patras Greece

ISBN: (纸本)9798400706356

Flexible Electronics (FE) offer distinct advantages, including mechanical flexibility and low process temperatures, enabling extremely low-cost production. To address the demands of applications such as smart sensors and wearables, flexible devices must be small and operate at low supply voltages. Additionally, target applications often require classifiers to operate directly on analog sensory input, necessitating the use of Analog to Digital Converters (ADCs) to process the sensory data. However, ADCs present serious challenges, particularly in terms of high area and power consumption, especially when considering stringent area and energy budget. In this work, we target common classifiers in this domain such as MLPs and SVMs and present a holistic approach to mitigate the elevated overhead of analog to digital interfacing in FE. First, we propose a novel design for Binary Search ADC that reduces area overhead 2× compared with the state-of-the-art Binary design and up to 5.4× compared with Flash ADC. Next, we present an in-training ADC optimization in which we keep the bare-minimum representations required and simplifying ADCs by removing unnecessary components. Our in-training optimization further reduces on average the area in terms of transistor count of the required ADCs by 5× for less than 1% accuracy loss. © 2025 Institute of Electrical and Electronics Engineers Inc.. All rights reserved.

关键词： Budget control

来源：评论

学校读者我要写书评

暂无评论

Personalizing Federated Instrument Segmentation With Visual Trait Priors in Robotic Surgery

引用

IEEE Transactions on Biomedical engineering 2025年第6期72卷 1886-1896页

作者： Xu, Jialang Wang, Jiacheng Yu, Lequan Stoyanov, Danail Jin, Yueming Mazomenos, Evangelos B. University College London UCL Hawkes Institute The Department of Medical Physics and Biomedical Engineering London United Kingdom Xiamen University Department of Computer Science School of Informatics Xiamen China The University of Hong Kong Department of Statistics and Actuarial Science Hong Kong National University of Singapore Department of Biomedical Engineering The Department of Electrical and Computer Engineering Singapore

Personalized federated learning (PFL) for surgical instrument segmentation (SIS) is a promising approach. It enables multiple clinical sites to collaboratively train a series of models in privacy, with each model tailored to the individual distribution of each site. Existing PFL methods rarely consider the personalization of multi-headed self-attention, and do not account for appearance diversity and instrument shape similarity, both inherent in surgical scenes. We thus propose PFedSIS, a novel PFL method with visual trait priors for SIS, incorporating global-personalized disentanglement (GPD), appearance-regulation personalized enhancement (APE), and shape-similarity global enhancement (SGE), to boost SIS performance in each site. GPD represents the first attempt at head- wise assignment for multi-headed self-attention personalization. To preserve the unique appearance representation of each site and gradually leverage the inter-site difference, APE introduces appearance regulation and provides customized layer- wise aggregation solutions via hypernetworks for each site's personalized parameters. The mutual shape information of instruments is maintained and shared via SGE, which enhances the cross-style shape consistency on the image level and computes the shape-similarity contribution of each site on the prediction level for updating the global parameters. PFedSIS outperforms state-of-the-art methods with +1.51% Dice, +2.11% IoU, -2.79 ASSD, -15.55 HD95 performance gains. © 1964-2012 IEEE.

关键词： Robotic surgery

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：