检索结果-内蒙古大学图书馆

Training Latency Minimization for Model-Splitting Allowed Federated Edge Learning

IEEE Transactions on Network science and engineering 2025年第3期12卷 2081-2092页

作者： Wen, Yao Zhang, Guopeng Wang, Kezhi Yang, Kun China University of Mining and Technology School of Computer Science and Technology Xuzhou221116 China Brunel University London Department of Computer Science MiddlesexUB8 3PH United Kingdom Nanjing University State Key Laboratory of Novel Software Technology Nanjing210008 China School of Intelligent Software and Engineering Nanjing University Suzhou Campus Suzhou215163 China

To alleviate the shortage of computing power faced by clients in training deep neural networks (DNNs) using federated learning (FL), we leverage the edge computing and split learning to propose a model-splitting allowed FL (SFL) framework, with the aim to minimize the training latency without loss of test accuracy. Under the synchronized global update setting, the latency to complete a round of global training is determined by the maximum latency for the clients to complete a local training session. Therefore, the training latency minimization problem (TLMP) is modelled as a minimizing-maximum problem. To solve this mixed integer nonlinear programming problem, we first propose a regression method to fit the quantitative-relationship between the cut-layer and other parameters of an AI-model, and thus, transform the TLMP into a continuous problem. Considering that the two subproblems involved in the TLMP, namely, the cut-layer selection problem for the clients and the computing resource allocation problem for the parameter-server are relative independence, an alternate-optimization-based algorithm with polynomial time complexity is developed to obtain a high-quality solution to the TLMP. Extensive experiments are performed on a popular DNN-model EfficientNetV2 using dataset MNIST, and the results verify the validity and improved performance of the proposed SFL framework. © 2013 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

MLPFormer: MLP-integrated transformer for colorectal histopathology whole slide image segmentation

引用

Neural Computing and Applications 2025年第6期37卷 4651-4661页

作者： Wang, Yuxuan Li, Dan Li, Xuechen Guo, Yan Zuo, Yanfei Shen, Linlin Department of Computer Science Wenzhou-Kean University Wenzhou China School of Electronics and Information Engineering Wuyi University Jiangmen China Department of Histo Pathology Shanghai Histo Pathology Diagnostic Center Shanghai China College of Computer Sicence and Software Engineering Shenzhen University Shenzhen China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University Shenzhen China

Colorectal intraepithelial neoplasia is a precancerous lesion of colorectal cancer, which is mainly diagnosed using pathological images. According to the characteristics of lesions, precancerous lesions can be classified into four different grades, i.e., background, normal, low and high level. However, due to the continuity of lesion development, tissue in different stages has high similarity. Recently, visual transformers have achieved impressive results in different visual tasks, due to their capacity of capturing long-range dependencies. However, due to computational cost, transformer cannot well extract detailed feature, which is important for lesion grading. To accurately segment pathological tissues with high similarity, in this work, we embed the multi-head MLP mixer module in the transformer network to extract cell and edge information. Furthermore, we propose a hierarchical MLP decoder to facilitate multi-scale feature fusion. The proposed transformer, namely MLPFormer, achieves remarkable results against competitive baselines on the Histo-CRC Biopsy dataset. The experimental results demonstrate that MLPFormer significantly outperforms the competitive baselines, i.e., a 3% dice improvement is achieved over the SegFormer series. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Cost-Optimized Task Offloading for Dependent Applications in Collaborative Edge and Cloud Computing

引用

IEEE Internet of Things Journal 2025年第9期12卷 12975-12988页

作者： Yuan, Haitao Hu, Qinglong Wang, Shen Bi, Jing Buyya, Rajkumar Lu, Jinhu Yang, Jinhong Zhang, Jia Zhou, MengChu Beihang University School of Automation Science and Electrical Engineering Beijing100191 China City University of Hong Kong Department of Computer Science Kowloon Tong Hong Kong Beijing University of Technology School of Software Engineering Faculty of Information Technology Beijing100124 China Laboratory School of Computing and Information Systems MelbourneVIC3010 Australia CSSC Systems Engineering Research Institute Beijing100036 China Southern Methodist University Lyle School of Engineering Department of Computer Science DallasTX75205 United States New Jersey Institute of Technology Department of Electrical and Computer Engineering NewarkNJ07102 United States

A collaborative system that includes mobile devices (MDs), edge nodes (ENs), and the cloud is needed where ENs at the network edge can run offloaded tasks of MDs with limited resources and energy for timely processing for latency-sensitive applications. Unlike existing studies, we formulate a total cost minimization problem for the system for applications, which can be divided into several interdependent subtasks. Each subtask can be executed in MDs, ENs, and the cloud. This work formulates a mixed-integer nonlinear program to minimize the total system cost. To address it, a novel meta-heuristic optimization algorithm called Genetic Simulated-annealing-based Particle swarm optimization with Auto-Encoder (GSPAE) is proposed, which innovatively combines feature extraction of deep learning and global search of meta-heuristic optimization. Genetic operations provide diverse solutions, the Metropolis acceptance of annealing offers a robust global search, and autoencoders (AEs) extract distribution characteristics of particles toward high-quality regions for fast convergence. Thus, GSPAE optimizes the associations between ENs and MDs and the scheduling of subtasks among MDs, ENs, and the cloud. Experiments with large-scale Google cluster datasets show that compared to state-of-the-art benchmark methods, GSPAE reduces the total cost by at least 17% while strictly meeting limits of application latency, available energy, computing, and communication resources of ENs and MDs. © 2014 IEEE.

关键词： Mobile edge computing

来源：评论

学校读者我要写书评

暂无评论

Ensemble Classifier for EEG-Based Stress Classification: An Empirical Study on Stacking Classifiers 3rd

Ensemble Classifier for EEG-Based Stress Classification: An ...

引用

3rd International Conference on Computational Electronics for Wireless Communications, ICCWC 2023

作者： Shikha, Shikha Sethia, Divyashikha Indu, S. Department of Computer Science and Engineering Delhi Technological University Delhi India Department of Software Engineering Delhi Technological University Delhi India Department of Electronics and Communication Delhi Technological University Delhi India

ISBN: (纸本)9789819719426

In the last few years, combining multiple algorithms to improve the performance of machine learning models has been a common practice. However, its application to stress detection still needs to be explored. This paper uses a stacking ensemble technique to introduce a novel stress classification approach using electroencephalogram (EEG) signals. Specifically, the study evaluates the effectiveness of various classifiers, including Random Forest (RF), K-Nearest Neighbor (KNN), Gradient Boosting (GB), Naive Bayes (NB), Support Vector Machine (SVM), Logistic Regression (LR), Decision Tree (DT), and Adaboost (AB), individually and in combination with RF as the base model, for stress classification tasks. The paper uses the SAM40 Dataset consisting of 32-channel EEG data to extract features in time and frequency domains for stress classification. The results show that the stacked classifiers outperform single classifiers, with RF + KNN providing the highest accuracy of 98.55%. The findings suggest that a stacked classifier is a promising approach for stress classification, as it can leverage the strengths of different algorithms and improve generalization performance. It holds promising future applicability in personalized stress management, healthcare interventions, and addressing societal concerns related to mental well-being. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

Fair and Communication-Efficient Personalized Federated Learning 18th

Fair and Communication-Efficient Personalized Federated Lea...

引用

18th International Conference on Wireless Artificial Intelligent Computing Systems and Applications, WASA 2024

作者： Zheng, Yifu Zheng, Xu Wang, Tingqi Mu, Chong Zhakiyev, Nurkhat School of Computer Science and Engineering University of Electronic Science and Technology of China Chengdu China School of Information and Software Engineering University of Electronic Science and Technology of China Chengdu China Department of Science and Innovation Astana IT University Astana Kazakhstan

ISBN: (纸本)9783031714665

Personalized Federated Learning (pFL) is among the most popular tasks in distributed deep learning, which compensates for mutual knowledge and enables device-specific model personalization. However, the effectiveness of pFL is severely impeded by challenges on fairness and significant communication overhead, where devices holding essential samples have to devote extensive resources for model training. To address these issues, we introduce the Fair and Communication-Efficient Personalized Federated Learning (FCE-PFL) framework, which harmonizes performance and device fairness while maintaining communication costs. Based on the Dempster-Shafer Theory, FCE-PFL employs assistance and contribution metrics to quantify the auxiliary information a client receives and provides. Then FCE-PFL balances and adjusts the involvement of devices in each training iteration, so as to facilitate fair training by constraining the maximum resource consumption per device and reduce the overall communication overhead. Our framework has been proven to be superior to existing methods in accuracy through rigorous experiments on the CIFAR10 and CIFAR100 datasets, demonstrating its potential as a fair and efficient pFL solution. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

RandoMix: a mixed sample data augmentation method with multiple mixed modes

引用

Multimedia Tools and Applications 2025年第8期84卷 4343-4359页

作者： Liu, Xiaoliang Shen, Furao Zhao, Jian Nie, Changhai National Key Laboratory for Novel Software Technology Nanjing University Nanjing China Department of Computer Science and Technology Nanjing University Nanjing China School of Artificial Intelligence Nanjing University Nanjing China School of Electronic Science and Engineering Nanjing University Nanjing China

Data augmentation plays a crucial role in enhancing the robustness and performance of machine learning models across various domains. In this study, we introduce a novel mixed-sample data augmentation method called RandoMix. RandoMix is specifically designed to simultaneously address robustness and diversity challenges. It leverages a combination of linear and mask mixed modes, introducing flexibility in candidate selection and weight adjustments. We evaluate the effectiveness of RandoMix on diverse datasets, including CIFAR-10/100, Tiny-ImageNet, ImageNet, and Google Speech Commands. Our results demonstrate its superior erformance compared to existing techniques such as Mixup, CutMix, Fmix, and ResizeMix. Notably, RandoMix excels in enhancing model robustness against adversarial noise, natural noise, and sample occlusion. The comprehensive experimental results and insights into parameter tuning underscore the potential of RandoMix as a versatile and effective data augmentation method. Moreover, it seamlessly integrates into the training pipeline. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Data augmentation Diversity Mixed-sample RandoMix Robustness

来源：评论

学校读者我要写书评

暂无评论

A novel deep learning approach for investigating liquid fuel injection in combustion system

引用

Discover Artificial Intelligence 2025年第1期5卷 1-14页

作者： Inam, Syed Azeem Khan, Abdullah Ayub Ahmed, Noor Mazhar, Tehseen Shahzad, Tariq Khan, Sunawar Saeed, Mamoon M. Hamam, Habib Department of Artificial Intelligence and Mathematical Science Sindh Madressatul Islam University Sindh Karachi Pakistan Department of Computer Science and Information Technology Shaheed Benazir Bhutto University Lyari Karachi Pakistan Department of Computer Science Shaheed Zulfiqar Ali Bhutto Institute of Science and Technology Karachi Pakistan Department of Computer Science School Education Department Government of Punjab Layyah31200 Pakistan School of Computer Science National College of Business Administration and Economics Lahore54000 Pakistan Department of Computer Engineering COMSATS University Islamabad Sahiwal Campus Sahiwal57000 Pakistan Department of Software Engineering Islamia University of Bahawalpur Bahawalnagar62300 Pakistan Sana’a00967 Yemen Faculty of Engineering Uni de Moncton MonctonNBE1A3E9 Canada Department of Electrical and Electronic Engineering Sciences University of Johannesburg Johannesburg South Africa Av. Grandes Ecoles BP 1989 Libreville Gabon Bridges for Academic Excellence Spectrum Center-ville Tunis Tunisia

The intricacies and instability of introducing cryogenic propellants into the combustion system have piqued the curiosity of scientists studying the procedure. The latest innovation is utilizing data-driven machine learning and deep learning approaches to gain deeper insights into the related difficulties. However, the current work serves as a baseline for future research because relatively few studies have used data-driven methodologies to assess the temperature of liquid fuel injections in combustion systems. The performance of Linear Regression (LR), Random Forest (RF), Extra Trees Regressor (ETR), Polynomial Regression (PR), Support Vector Regressor (SVR), Decision Tree Regressor (DTR), Gradient Boost Regressor (GBR), XGB Regressor (XGBoost), AdaBoost Regressor (ABR), K-Neighbors Regressor (KNR), Long-Short Term Memory (LSTM), Bi-LSTM (Bi-directional Long-Short Term Memory) has all been investigated in this study. The study also suggested a Fully Connected Neural Network (FCNN) to examine its performance and paired it with an Extra Tree Regressor (ETR). The coupled FCNN and Extra Tree Regressor outperform the other algorithms with a Mean Square Error (MSE) of 0.0000005062, Root Mean Square Error (RMSE) of 0.00071148, Mean Absolute Error (MAE) of 0.00020672, and R-squared (R2) value of 0.99998689. Linear Regression, Polynomial Regression, and Support Vector Regressor are found to be the least-performing algorithms. The current work uses machine learning and deep learning methods to make data-driven decisions for liquid fuel injection in the combustion system. © The Author(s) 2025.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

A modified grey wolf optimizer with multi-solution crossover integration algorithm for feature selection

引用

Journal of Ambient Intelligence and Humanized Computing 2025年第1期16卷 329-345页

作者： Ihsan, Muhammad Din, Fakhrud Zamli, Kamal Z. Ghadi, Yazeed Yasin Alahmadi, Tahani Jaser Innab, Nisreen Faculty of IT Department of Computer Science and IT University of Malakand Khyber Pakhtunkhwa Chakdara18800 Pakistan Faculty of Computing College of Computing and Applied Sciences Universiti Malaysia Pahang Pahang Pekan26600 Malaysia Department of computer science and software engineering Al Ain University Abu Dhabi12555 United Arab Emirates Department of Information SystemsCollege of Computer and Information Sciences Princess Nourah bint Abdulrahman University Riyadh84428 Saudi Arabia Department of Computer Science and Information SystemsCollege of Applied Sciences Al-Maarefa University Diriyah Riyadh13713 Saudi Arabia

Feature selection helps eradicate redundant features which is essential to mitigate the curse of dimensionality when a machine-learning model deals with high-dimensional datasets. Grey Wolf Optimizer (GWO) is a swarm-based algorithm that simulates the wolves’ hunting behavior. Although very efficient, GWO faces some limitations which may cause premature convergence and/or local optima trapping. Moreover, GWO relies mainly on the three best wolves, limiting its potential for diverse exploration and exploitation. This work proposes an improved version of GWO namely, a modified grey wolf optimizer with multi-solution crossover integration (MGWO-MCI) algorithm. MGWO-MCI algorithm incorporates a multi-solution strategy that evolves new potential solutions in the optimization process. A crossover operation is performed between the new wolves and the existing hierarchy, reforming the position-updating process. MGWO-MCI utilizes this position-updating process using two different approaches. The first approach named MGWO-MCI-I expands the additional wolves’ role to both exploration and exploitation whereas the second approach named MGWO-MCI-II incorporates their role to exploration only. These approaches are evaluated and tested using 18 datasets and an Intrusion detection dataset NSL-KDD for feature selection. Statistically, the results are analyzed through the Wilcoxon test, which shows the superiority of MGWO-MCI-II. MGWO-MCI-II outperforms others with an accuracy of 98.6% on NSL-KDD and achieves 55.5% overall best outcomes on other datasets. Moreover, the MGWO-MCI was evaluated on two constrained optimization problems, the pressure vessel and welded beam design validating its effectiveness and adaptability in solving different optimization problems. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025.

关键词： Constrained optimization

来源：评论

学校读者我要写书评

暂无评论

Optimizing Secure Multi-User ISAC Systems With STAR-RIS: A Deep Reinforcement Learning Approach for 6G Networks

引用

IEEE Access 2025年 13卷 31472-31484页

作者： Kamal, Mian Muhammad Zain Ul Abideen, Syed Al-Khasawneh, M.A. Alabrah, Amerah Sohail Ahmed Larik, Raja Irfan Marwat, Muhammad Southeast University School of Electronic Science and Engineering Jiangning Jiangsu Nanjing211189 China Qingdao University College of Computer Science and Technology Qingdao266071 China Al-Ahliyya Amman University Hourani Center for Applied Scientific Research Amman19111 Jordan Skyline University College School of Computing University City Sharjah Sharjah United Arab Emirates King Saud University College of Computer and Information Science Department of Information Systems Riyadh11543 Saudi Arabia Ilma University Department of Computer Science Sindh Karachi75190 Pakistan University of Science and Technology Bannu Department of Software Engineering Bannu28100 Pakistan

The rapid evolution of wireless communication technologies and the increasing demand for multi-functional systems have led to the emergence of integrated sensing and communication (ISAC) as a key enabler for future 6G networks. Simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RISs) have recently garnered significant attention for their ability to enhance signal coverage and improve system efficiency. This paper investigates a STAR-RIS-assisted ISAC system designed to secure communication for multiple legitimate users (LUs) while safeguarding against multiple eavesdroppers (Eves). By jointly optimizing the base station (BS) transmit beamforming, STAR-RIS transmission and reflection coefficients, and receive filters, the proposed framework aims to maximize the long-term average secrecy rate for all LUs. Constraints are imposed to ensure minimum echo signal-to-noise ratios (SNRs) for sensing and meet the achievable rate requirements of LUs. To address the inherent complexity of this non-convex problem, two deep reinforcement learning (DRL) algorithms are proposed. Numerical results demonstrate that the proposed system achieves significant improvements in secrecy rate compared to conventional RIS setups. This work provides a scalable and efficient approach for secure multi-user ISAC systems, making it highly relevant for future 6G networks, smart cities, and IoT applications. © 2025 The Authors.

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

DIG: Improved DINO for Graffiti Detection

引用

IEEE Transactions on Systems, Man, and Cybernetics: Systems 2025年第5期55卷 3557-3569页

作者： Wang, Bingshu Mao, Qianchen Liu, Aifei Chen, Long Chen, C. L. Philip Northwestern Polytechnical University School of Software Xi’an710129 China Shenzhen University Guangdong Key Laboratory of Intelligent Information Processing Shenzhen Key Laboratory of Media Security Shenzhen518060 China University of Macao Department of Computer and Information Science China South China University of Technology School of Computer Science and Engineering Guangzhou510641 China

Graffiti detection is essential in historic building protection and urban neighborhood management. Graffiti detection has made significant progress in recent years based on the development of deep learning. However, small-scale graffiti, interference from the background, and the false detection of word parts in graffiti make it a challenging problem. This article proposes a Transformer-based high-precision graffiti detection method, namely DIG. Precisely, it consists of three modules: 1) Spatial query selection (SQS), scale-aware IoU loss (SIL);2) Denoising Task with binary contrastive denoising (BCDN);and 3) IoU-guided box denoising (IBD) modules. To detect small-scale graffiti, this paper proposes SIL to help the loss function to perceive small-scale graffiti and large-scale graffiti fairly. To reduce the false detection of word parts, this article presents the SQS module, which integrates spatial information into the query selection process of the Encoder to filter out falsely detected bounding boxes within the graffiti. To reduce the interference from the background, this article introduces a denoising task with BCDN and IBD modules, improving the model’s ability to distinguish graffiti from the background and accurately select appropriate bounding boxes. A large number of experimental results on the STORM dataset show that our method achieves state-of-the-art results with an AP50 of 87.9%. Moreover, DIG achieved competitive results on the FineFM dataset for mask detection. This indicates that DIG can also be conveniently transferred to detect other scenarios. © 2013 IEEE.

关键词： Historic preservation

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：