检索结果-内蒙古大学图书馆

On-device diagnostic recommendation with heterogeneous federated BlockNets

Science China(information Sciences) 2025年第4期68卷 33-49页

作者： Minh Hieu NGUYEN Thanh Trung HUYNH Thanh Toan NGUYEN Phi Le NGUYEN Hien Thu PHAM Jun JO Thanh Tam NGUYEN School of Information and Communication Technology Griffith University School of Computer and Communication Sciences Ecole Polytechnique Federale de Lausanne Faculty of Information Technology HUTECH University School of Information and Communication Technology Hanoi University of Science and Technology Commonwealth Scientific and Industrial Research Organization (CSIRO)

The evolution of edge computing has advanced the accessibility of E-health recommendation services, encompassing areas such as medical consultations, prescription guidance, and diagnostic assessments. Traditional methodologies predominantly utilize centralized recommendations, relying on servers to store client data and dispatch advice to ***, these conventional approaches raise significant concerns regarding data privacy and often result in computational inefficiencies. E-health recommendation services, distinct from other recommendation domains, demand not only precise and swift analyses but also a stringent adherence to privacy safeguards, given the users' reluctance to disclose their identities or health information. In response to these challenges, we explore a new paradigm called on-device recommendation tailored to E-health diagnostics, where diagnostic support(such as biomedical image diagnostics), is computed at the client *** leverage the advances of federated learning to deploy deep learning models capable of delivering expert-level diagnostic suggestions on clients. However, existing federated learning frameworks often deploy a singular model across all edge devices, overlooking their heterogeneous computational capabilities. In this work, we propose an adaptive federated learning framework utilizing BlockNets, a modular design rooted in the layers of deep neural networks, for diagnostic recommendation across heterogeneous devices. Our framework offers the flexibility for users to adjust local model configurations according to their device's computational power. To further handle the capacity skewness of edge devices, we develop a data-free knowledge distillation mechanism to ensure synchronized parameters of local models with the global model, enhancing the overall accuracy. Through comprehensive experiments across five real-world datasets, against six baseline models, within six experimental setups, and various data distribution scenario

关键词： intelligent recommendation federated learning heterogeneous devices E-health diagnostics

来源：评论

学校读者我要写书评

暂无评论

Classification of cervical cells from the Pap smear image using the RES_DCGAN data augmentation and ResNet50V2 with self-attention architecture

引用

Neural Computing and Applications 2024年第34期36卷 21801-21815页

作者： Wubineh, Betelhem Zewdu Rusiecki, Andrzej Halawa, Krzysztof Faculty of Information and Communication Technology Wroclaw University of Science and Technology Wroclaw Poland

Cervical cancer is a type of cancer in which abnormal cell growth occurs on the surface lining of the cervix. In this study, we propose a novel residual deep convolutional generative adversarial network (RES_DCGAN) for data augmentation and ResNet50V2 self-attention method to classify cervical cells, to improve the generalizability and performance of the model. The proposed method involves adding residual blocks in the generator of the DCGAN to enhance data flow and generate higher-quality images. Subsequently, a self-attention mechanism is incorporated at the top of the pre-trained models to allow the model to focus more on significant features of the input data. To evaluate our approach, we utilized the Pomeranian and SIPaKMeD cervical cell imaging datasets. The results demonstrate superior performance, achieving an accuracy of 98% with Xception and 96.4% with ResNet50V2 on the Pomeranian dataset. Additionally, DenseNet121 with self-attention achieved accuracies of 92% and 95% in multiclass and binary classification, respectively, using the SIPaKMeD dataset. In conclusion, our RES_DCGAN-based data augmentation and pre-trained with self-attention model yields a promising result in the classification of cervical cancer cells. © The Author(s) 2024.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Privacy-preserving explainable AI: a survey

引用

Science China(information Sciences) 2025年第1期68卷 23-56页

作者： Thanh Tam NGUYEN Thanh Trung HUYNH Zhao REN Thanh Toan NGUYEN Phi Le NGUYEN Hongzhi YIN Quoc Viet Hung NGUYEN School of Information and Communication Technology Griffith University School of Computer and Communication Sciences Ecole Polytechnique Federale de Lausanne Faculty of Mathematics and Computer Science University of Bremen Faculty of Information Technology HUTECH University Department of Computer Science Hanoi University of Science and Technology School of Electrical Engineering and Computer Science The University of Queensland

As the adoption of explainable AI(XAI) continues to expand, the urgency to address its privacy implications intensifies. Despite a growing corpus of research in AI privacy and explainability, there is little attention on privacy-preserving model explanations. This article presents the first thorough survey about privacy attacks on model explanations and their countermeasures. Our contribution to this field comprises a thorough analysis of research papers with a connected taxonomy that facilitates the categorization of privacy attacks and countermeasures based on the targeted explanations. This work also includes an initial investigation into the causes of privacy leaks. Finally, we discuss unresolved issues and prospective research directions uncovered in our analysis. This survey aims to be a valuable resource for the research community and offers clear insights for those new to this domain. To support ongoing research, we have established an online resource repository, which will be continuously updated with new and relevant findings.

关键词： privacy-preserving explainable AI privacy attacks privacy defences PrivEx PPXAI

来源：评论

学校读者我要写书评

暂无评论

Effect of Data Division on Classification Performance of Overlapping Red Blood Cells using Sharpening Image Segmentation Strategy and Convolutional Neural Network

IAENG International Journal of Computer Science

引用

IAENG International Journal of Computer Science 2024年第3期51卷 204-211页

作者： Arianti, Nunik Destria Ahmad, Norashikin Muda, Azah Kamilah Faculty of Information and Communication Technology Universiti Teknikal Malaysia Melaka Durian TunggalMelaka76100 Malaysia Faculty of Information and Communication Technology Universiti Teknikal Malaysia Melaka Durian TunggalMelaka76100 Malaysia

Red blood cell (RBC) counts and identification must be done accurately to diagnose various illnesses. Although there are drawbacks to automated image analysis methods compared to hand counting, they include differences in picture characteristics and overlapped cells. This study aimed to create a robust model that can classify overlapping RBCs by employing a region-based convolution neural network (R-CNN) with several data divisions and a sharpening picture segmentation technique. Three sharpening approaches, wand, clahe, and Otsu threshold, were incorporated into the assemble boosting model utilizing 100 RBC image datasets from ***. The integrated strategy sought to improve image segmentation quality before employing R-CNN for classification. The model's effectiveness was evaluated at many data split percentages, ranging from 80/20 to 60/40. Across many data splits, the model's average accuracy standard was 0.82 ± 0.09. It performed best when the data split was 75/25. Statistical analysis confirmed Significant performance differences between these groups, and overlapping RBC categorization caused more significant difficulties than single RBCs. The model's robustness was demonstrated by its performance, which was resistant to variations in data splits. RBCs were successfully categorized using the combined sharpening and R-CNN technique, especially single cells. Overlapping RBC categorization is still tricky, highlighting the need for more study in this field. Future advancements in biological image analysis have a promising precedent thanks to the model's stability across different data splits. © (2024), (International Association of Engineers). All Rights Reserved.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Central Tendency Feature Selection (CTFS): a novel approach for efficient and effective feature selection in intrusion detection systems

引用

Multimedia Tools and Applications 2025年 1-22页

作者： Bouke, Mohamed Aly Abdullah, Azizol Udzir, Nur Izura Samian, Normalia Department of Communication Technology and Network Faculty of Computer Science and Information Technology Universiti Putra Malaysia 43400 Serdang Malaysia

In the digital era, the escalation of data generation and cyber threats has heightened the importance of network security. Machine Learning-based Intrusion Detection Systems (IDS) play a crucial role in combating these threats, yet there is a significant gap in their feature selection methodologies. These traditional methods often fail to handle large datasets, leading to performance limitations efficiently. To advance this crucial aspect of IDS, we propose the novel Central Tendency-Based Feature Selection (CTFS) method. Our empirical investigation, utilizing the widely acknowledged UNSWNB15 and NSLKDD cybersecurity datasets, reveals that CTFS not only outperforms established feature selection methods like Univariate Selection, Recursive Feature Elimination (RFE), Principal Component Analysis (PCA), and Gini Index in terms of accuracy, achieving up to 95%, and recall, reaching up to 99%, but it also substantially enhances execution efficiency. Specifically, CTFS demonstrates a remarkable reduction in feature selection time to approximately 0.1 s for UNSWNB15 and 0.026 s for NSLKDD, outmatching the more time-intensive RFE, which recorded 569.7 s and 24.9 s, respectively. These findings showcase CTFS as a potent solution, balancing high performance with computational efficiency, a key consideration in time-sensitive security applications. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2025.

关键词： Cybersecurity

来源：评论

学校读者我要写书评

暂无评论

An adaptive fuzzy weight algorithm for the class imbalance learning problem

引用

International Journal of Intelligent information and Database Systems 2024年第3期16卷 221-240页

作者： Quang, Vo Duc Khang, Tran Dinh School of Information and Communication Technology Hanoi University of Science and Technology Hanoi Viet Nam Faculty of Information Technology Vinh University Nghean Viet Nam

In this study, we propose an adaptive fuzzy weight algorithm for the problem of two-class imbalanced learning. Initially, our algorithm finds a set of fuzzy weight values for data samples based on the distance from each sample to the centres of both minority and majority classes. Then, our algorithm iteratively adjusts the fuzzy weight values of sensitive samples on either positive or negative margins or class label noises. By doing so, our algorithm increases the influence of minority samples and decreases the influence of majority samples in forming a classifier model. Experimental results on four benchmark real-world imbalanced datasets including Transfusion, Ecoli, Yeast, and Abalone show that our algorithm outperforms the fuzzy SVM-CIL algorithm in terms of classification performance. © 2024 Inderscience Enterprises Ltd.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Joint user association and resource allocation for cost-efficient NOMA-enabled F-RANs

引用

Digital communications and Networks 2024年第6期10卷 1686-1697页

作者： Yuan Ai Chenxi Liu Mugen Peng Faculty of Information Technology Beijing University of TechnologyBeijing 100124China SchoolofInformation and Communication Enginering Beijing University of Posts and TelecommunicationsBeijing 100876China

Integrating Non-Orthogonal Multiple Access(NOMA)into Fog Radio Access Networks(F-RANs)has shown to be effective in boosting the spectral efficiency,energy efficiency,connectivity,and reducing the latency,thus attracting significant research ***,the performance improvement of the NOMA-enabled F-RANs is at the cost of computational overheads,which are commonly neglected in their design and *** address this issue,in this paper,we propose a hybrid dynamic downlink framework for NOMA-enabled *** this framework,we first develop a novel network utility function,which takes both the network throughput and computational overheads into consideration,thus enabling us to comprehensively evaluate the performance of different access schemes for *** on the developed network utility function,we further formulate a network utility maximization problem,subject to practical constraints on the decoding order,power allocation,and *** solve this NP-hard problem,we decompose it into two subproblems,namely,a user equipment association and subchannel assignment subproblem and a power allocation *** matching and sequential convex programming-based algorithms are designed to solve these two subproblems,*** numerical results,we show how our proposed algorithms can achieve a good balance between the network throughput and computational overheads by judiciously adjusting the maximum transmit power of fog access *** also show that the proposed NOMA-enabled F-RAN framework can increase,by up to 89%,the network utility,compared to OMA-based F-RANs.

关键词： Non-orthogonal multiple access(NOMA) Resource allocation Fog radio access networks

来源：评论

学校读者我要写书评

暂无评论

Dopamine Functionalized Activated Carbon-Based 3-D-Printed Biodegradable Electrode for Supercapacitor

IEEE Journal on Flexible Electronics

引用

IEEE Journal on Flexible Electronics 2025年第1期4卷 42-51页

作者： Mevada, Chirag Parihar, Vijay Singh Tewari, Amit Keskinen, Jari Kellomäki, Minna Mäntysalo, Matti Faculty of Information Technology and Communication Sciences Tampere University Tampere33720 Finland Faculty of Medicine and Health Technology Tampere University Tampere33720 Finland

This study investigates dopamine-grafted activated carbon (DAC) composites synthesized via chemical bonding for use as biodegradable supercapacitor (SC) electrodes. This composite approach leverages both electrical double-layer capacitance and pseudocapacitance to achieve a significant increase in capacitance. The results of thermogravimetric, surface area and porosimeter, Raman, and X-ray photoelectron spectroscopy (XPS) analysis confirmed the grafting of dopamine (DP) on activated carbon (AC). The 3-D-printed SCs were manufactured utilizing electrode materials based on DAC and AC to investigate their electrochemical performance. The DAC-based SC exhibited a specific capacitance of 24 F·g−1, notably higher than the 15-F·g−1 specific capacitance observed in AC-based SC under a current density of 0.008 A·g−1. This difference highlights the important role played by the redox reaction facilitated by the grafted dopamine molecule. DAC holds promise as a biodegradable electrode material suitable for SCs for low-power wireless sensor application. © 2022 IEEE.

关键词： Redox reactions

来源：评论

学校读者我要写书评

暂无评论

LISTENING IN COURT-CONNECTED MEDIATIONS

引用

International Journal of Listening 2024年第3期38卷 186-202页

作者： Ala-Kortesmaa, Sanna Valikoski, Tuula-Riitta Faculty of Information Technology and Communication Sciences Tampere University Finland

As the use of court-connected mediations has received a permanent status as an intrajudicial conflict resolution method, the purpose of this study (n = 38/44 Finnish judges) was to elucidate how listening is used in mediations. The data set one consisted of answers to open-ended questions and was analyzed using thematic content analysis. The data set two consisted 56 statements that were used to create a personal listening profile for each participating judge. The results suggest that 1) facilitative, evaluative, and transformative mediation models have both shared and different listening goals that require an extensive listening competence from the mediator judges, 2) complexity and relationality characterize listening in facilitative mediations, and 3) most judges sent highly or moderately person-centered messages and were excellent in interpretation and evaluation. These results suggest that the relationality of listening in court-connected mediations is imperative in meeting the needs of the mediating parties. The judges’ listening behavior, listening goals, and the listening requirements of each mediation style suggest that the full potential of listening in court-connected mediations is yet to be reached. © 2023 The Author(s). Published with license by Taylor & Francis Group, LLC.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Why are consumers dissatisfied? A text mining approach on Sri Lankan mobile banking apps

引用

International Journal of Intelligent Computing and Cybernetics 2023年第4期16卷 727-744页

作者： Maas Sherina Sally Department of Information and Communication Technology Faculty of TechnologyUniversity of ColomboColomboSri Lanka

Purpose:The motivation of this study is to identify whether the overall rating of a banking app actually reflects the customer opinion and to find the causes for reduced ***,these causes lead to the dissatisfaction of ***,these insights reflect the overall rating of the app and it is a source of information to the executive management to contemplate on their services and take timely and effective decisions to improve their mobile ***/methodology/approach:This research was conducted on ten reputed Sri Lankan mobile banking apps to analyze the textual opinions of the *** were collected from the Google Play Store considering the higher Android consumers in Sri *** review was automatically classified into a relevant sentiment(positive,negative or neutral).These classified reviews were examined along with its rating to identify any *** trends of the positive and negative reviews of each app were observed separately along with *** modeling techniques were used to identify the causes of such ***:Although banks expect to perpetuate good customer reviews all the time,there were aberrant negative trends observed during certain time *** results revealed that unstable versions after recent updates,bad customer service,erroneous functional and nonfunctional features are the root causes toward the dissatisfaction of the ***/value:No previous study has been done on the textual reviews of Sri Lankan mobile banking *** studies had considered analyzing the reviews of the app on the entire period of its usage,whereas this research finds the trends where negative reviews surpass the positive reviews and analyze the causes of such behavior.

关键词： Topic modelling LDA Reviews and ratings Google Play Store Mobile banking

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：