检索结果-内蒙古大学图书馆

Shielding Federated Learning: Robust Aggregation with Adaptive Client Selection

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Wan, Wei Hu, Shengshan Lu, Jianrong Yu Zhang, Leo Jin, Hai He, Yuanyuan School of Cyber Science and Engineering Huazhong University of Science and Technology China School of Computer Science and Technology Huazhong University of Science and Technology China National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Hubei Engineering Research Center on Big Data Security China Cluster and Grid Computing Lab School of Information Technology Deakin University Australia

Federated learning (FL) enables multiple clients to collaboratively train an accurate global model while protecting clients’ data privacy. However, FL is susceptible to Byzantine attacks from malicious participants. Although the problem has gained significant attention, existing defenses have several flaws: the server irrationally chooses malicious clients for aggregation even after they have been detected in previous rounds;the defenses perform ineffectively against sybil attacks or in the heterogeneous data setting. To overcome these issues, we propose MAB-RFL, a new method for robust aggregation in FL. By modelling the client selection as an extended multi-armed bandit (MAB) problem, we propose an adaptive client selection strategy to choose honest clients that are more likely to contribute high-quality updates. We then propose two approaches to identify malicious updates from sybil and non-sybil attacks, based on which rewards for each client selection decision can be accurately evaluated to discourage malicious behaviors. MAB-RFL achieves a satisfying balance between exploration and exploitation on the potential benign clients. Extensive experimental results show that MAB-RFL outperforms existing defenses in three attack scenarios under different percentages of attackers. Copyright © 2022, The Authors. All rights reserved.

关键词： data privacy

Towards Efficient data-Centric Robust Machine Learning with Noise-based Augmentation

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Liu, Xiaogeng Wang, Haoyu Zhang, Yechao Wu, Fangzhou Hu, Shengshan School Of Cyber Science And Engineering National Engineering Research Center For Big Data Technology And System Services Computing Technology And System Lab Hubei Engineering Research Center On Big Data Security Huazhong University Of Science And Technology Wuhan430074 China

The data-centric machine learning aims to find effective ways to build appropriate datasets which can improve the performance of AI models. In this paper, we mainly focus on designing an efficient data-centric scheme to improve robustness for models towards unforeseen malicious inputs in the black-box test settings. Specifically, we introduce a noisedbased data augmentation method which is composed of Gaussian Noise, Salt-and-Pepper noise, and the PGD adversarial perturbations. The proposed method is built on lightweight algorithms and proved highly effective based on comprehensive evaluations, showing good efficiency on computation cost and robustness enhancement. In addition, we share our insights about the data-centric robust machine learning gained from our experiments. Copyright © 2022, The Authors. All rights reserved.

关键词： Gaussian noise (electronic)

Does Graph Distillation See Like Vision dataset Counterpart?

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Yang, Beining Wang, Kai Sun, Qingyun Ji, Cheng Fu, Xingcheng Tang, Hao You, Yang Li, Jianxin School of Computer Science and Engineering Beihang University China Advanced Innovation Center for Big Data and Brain Computing Beihang University China National University of Singapore Singapore Carnegie Mellon University United States

Training on large-scale graphs has achieved remarkable results in graph representation learning, but its cost and storage have attracted increasing concerns. Existing graph condensation methods primarily focus on optimizing the feature matrices of condensed graphs while overlooking the impact of the structure information from the original graphs. To investigate the impact of the structure information, we conduct analysis from the spectral domain and empirically identify substantial Laplacian Energy Distribution (LED) shifts in previous works. Such shifts lead to poor performance in cross-architecture generalization and specific tasks, including anomaly detection and link prediction. In this paper, we propose a novel Structure-broadcasting Graph dataset Distillation (SGDD) scheme for broadcasting the original structure information to the generation of the synthetic one, which explicitly prevents overlooking the original structure information. Theoretically, the synthetic graphs by SGDD are expected to have smaller LED shifts than previous works, leading to superior performance in both cross-architecture settings and specific tasks. We validate the proposed SGDD across 9 datasets and achieve state-of-the-art results on all of them: for example, on the YelpChi dataset, our approach maintains 98.6% test accuracy of training on the original graph dataset with 1,000 times saving on the scale of the graph. Moreover, we empirically evaluate there exist 17.6% ∼ 31.4% reductions in LED shift crossing 9 datasets. Extensive experiments and analysis verify the effectiveness and necessity of the proposed designs. The code is available in the https://***/RingBDStack/SGDD. Copyright © 2023, The Authors. All rights reserved.

关键词： Distillation

Model Inversion Attacks Through Target-Specific Conditional Diffusion Models

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Li, Ouxiang Hao, Yanbin Wang, Zhicai Zhu, Bin Wang, Shuo Zhang, Zaixi Feng, Fuli School of Artificial Intelligence and Data Science University of Science and Technology of China Hefei China School of Computer Science and Information Engineering Hefei University of Technology Hefei230601 China School of Computing and Information Systems Singapore Management University Singapore School of Information Science and Technology University of Science and Technology of China Hefei China Department of Computer Science and Engineering University of Science and Technology of China Hefei China

Model inversion attacks (MIAs) aim to reconstruct private images from a target classifier's training set, thereby raising privacy concerns in AI applications. Previous GAN-based MIAs tend to suffer from inferior generative fidelity due to GAN's inherent flaws and biased optimization within latent space. To alleviate these issues, leveraging on diffusion models' remarkable synthesis capabilities, we propose Diffusion-based Model Inversion (Diff-MI) attacks. Specifically, we introduce a novel target-specific conditional diffusion model (CDM) to purposely approximate target classifier's private distribution and achieve superior accuracy-fidelity balance. Our method involves a two-step learning paradigm. Step-1 incorporates the target classifier into the entire CDM learning under a pretrain-thenfinetune fashion, with creating pseudo-labels as model conditions in pretraining and adjusting specified layers with image predictions in fine-tuning. Step-2 presents an iterative image reconstruction method, further enhancing the attack performance through a combination of diffusion priors and target knowledge. Additionally, we propose an improved max-margin loss that replaces the hard max with top-k maxes, fully leveraging feature information and soft labels from the target classifier. Extensive experiments demonstrate that Diff-MI significantly improves generative fidelity with an average decrease of 20% in FID while maintaining competitive attack accuracy compared to state-of-the-art methods across various datasets and models. Our code is available at: https://***/Ouxiang-Li/Diff-MI. Copyright © 2024, The Authors. All rights reserved.

关键词： Differential privacy

Prognosis of Relocation Disease in Animals using Aggregation Method with Optimization Techniques

学校读者我要写书评

暂无评论

Prognosis of Relocation Disease in Animals using Aggregation...

International Conference on Parallel Distributed and Grid computing (PDGC)

作者： Khushwant Singh Mohit Yadav R K Yadav Bharat Bhushan Naib Sandeep Bhatia Deepak Kumar Panda Computer Science and Engineering University Institute of Engineering and Technology Maharshi Dayanand University Rohtak India Department of Mathematics University Institute of Sciences Chandigarh University Mohali India Department of Data Science Christ University Bengaluru India School of Computing Science and Engineering Galgotias University Greater Noida Uttar Pradesh India School of Computer Science and Engineering Galgotias University Greater Noida Uttar Pradesh India

ISBN: (数字)9798331521349

ISBN: (纸本)9798331521356

In the most cutting-edge setting, health data is processed by machine learning algorithms, which are used to forecast illnesses. Dementia, especially Alzheimer's disease, is a leading cause of diminished quality of life in the elderly. Early diagnosis by medical professionals increases the likelihood of reducing the aggressiveness of the disease. In this study, we develop a new uncertainty-based clustering model to handle the centroid selection ambiguity and the issue of noisy instances and outliers that lower the efficiency of prediction models. This work employs an uncertainty-based optimization technique to handle the unknown pattern of Alzheimer Disease (AD) patients, since it is relatively tough to handle unknown patterns with unsupervised learning algorithms. Converting the instances in the AD dataset to the membership value of the dependent variable allows for an accurate determination of whether they belong as AD patterns or non-AD patterns. This proposed study takes a migration-based optimization method to animal migration, where the best instances are chosen as centroids and fresh instances are evaluated for clustering; this minimizes outliers throughout the clustering process by using comparable patterns. To make sure, we check the fitness values of each instance; the ones with the highest values are called centroids. To control the unknowns when dealing with outliers, the fuzzy Euclidean distance is employed. By comparing it to current state-of-the-art clustering methods, the OASIS dataset simulation results show that the proposed uncertainty-based Animal Migration optimization method (UAMO) performs better.

关键词： Accuracy Uncertainty Animals Optimization methods Clustering algorithms Predictive models Prognostics and health management Alzheimer's disease Stress Unsupervised learning

Vision-Amplified Semantic Entropy for Hallucination Detection in Medical Visual Question Answering

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Liao, Zehui Hu, Shishuai Zou, Ke Fu, Huazhu Zhen, Liangli Xia, Yong National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology School of Computer Science and Engineering Northwestern Polytechnical University Xi’an710072 China Institute of High Performance Computing Agency for Science Technology and Research Singapore138632 Singapore National University of Singapore Singapore119077 Singapore

Multimodal large language models (MLLMs) have demonstrated significant potential in medical Visual Question Answering (VQA). Yet, they remain prone to hallucinations—incorrect responses that contradict input images, posing substantial risks in clinical decision-making. Detecting these hallucinations is essential for establishing trust in MLLMs among clinicians and patients, thereby enabling their real-world adoption. Current hallucination detection methods, especially semantic entropy (SE), have demonstrated promising hallucination detection capacity for LLMs. However, adapting SE to medical MLLMs by incorporating visual perturbations presents a dilemma. Weak perturbations preserve image content and ensure clinical validity, but may be overlooked by medical MLLMs, which tend to over rely on language priors. In contrast, strong perturbations can distort essential diagnostic features, compromising clinical interpretation. To address this issue, we propose Vision Amplified Semantic Entropy (VASE), which incorporates weak image transformations and amplifies the impact of visual input, to improve hallucination detection in medical VQA. We first estimate the semantic predictive distribution under weak visual transformations to preserve clinical validity, and then amplify visual influence by contrasting this distribution with that derived from a distorted image. The entropy of the resulting distribution is estimated as VASE. Experiments on two medical open-ended VQA datasets demonstrate that VASE consistently outperforms existing hallucination detection methods. Copyright © 2025, The Authors. All rights reserved.

关键词： Semantics

Motion Forecasting for Autonomous Vehicles: A Survey

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Shi, Jianxin Chen, Jinhao Wang, Yuandong Sun, Li Liu, Chunyang Xiong, Wei Wo, Tianyu Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University Beijing100191 China Information Engineering College Capital Normal University Beijing100048 China School of Control and Computer Engineering North China Electric Power University Beijing102206 China Didi Chuxing Beijing100094 China

In recent years, the field of autonomous driving has attracted increasingly significant public interest. Accurately forecasting the future behavior of various traffic participants is essential for the decision-making of Autonomous Vehicles (AVs). In this paper, we focus on both scenario-based and perception-based motion forecasting for AVs. We propose a formal problem formulation for motion forecasting and summarize the main challenges confronting this area of research. We also detail representative datasets and evaluation metrics pertinent to this field. Furthermore, this study classifies recent research into two main categories: supervised learning and self-supervised learning, reflecting the evolving paradigms in both scenario-based and perception-based motion forecasting. In the context of supervised learning, we thoroughly examine and analyze each key element of the methodology. For self-supervised learning, we summarize commonly adopted techniques. The paper concludes and discusses potential research directions, aiming to propel progress in this vital area of AV technology. © 2025, CC BY.

关键词： Self-supervised learning

Cost-Effective Communication in UDN in Indoor and Outdoor Environment via Machine Learning

学校读者我要写书评

暂无评论

Cost-Effective Communication in UDN in Indoor and Outdoor En...

2023 International Conference on Artificial Intelligence and Knowledge Discovery in Concurrent engineering, ICECONF 2023

作者： Karman, K. Nattar Velmurugan, V. Raju, Kommisetti Murthy Sajana, T. Vijayalakshmi, V. Dhanraj, JoshuvaArockia Department of Artificial Intelligence and Machine Learning Saveetha School of Engineering Tamil Nadu Chennai602105 India Department of Electronics and Communication Engineering Vel Tech Rangarajan and Dr.Sagunthala RandD Institute of Science and Technology Tamil Nadu Chennai600062 India Department of Electronics and Communication Engineering Shri Vishnu Engineering College for Women Andhra Pradesh West Godavari India Department of Artificial Intelligence and Data Science KoneruLakshmaiah Education Foundation Andhra Pradesh Vaddeswaram522502 India Department of Networking and Communications School of Computing SRM Institute of Science and Technology Tamil Nadu Kattankulathur603203 India Department of Mechatronics Engineering Hindustan Institute of Technology and Science Tamil Nadu Chennai603103 India

ISBN: (纸本)9798350334357

In general, applications on a densely populated network are slower. When there are no opportunities to interact with the devices on the network, the user is forced to communicate at some cost. Thus, the inconsistency that exists there increases the cost over that dense network. The single way to evaluate prospect usage is to look at that have used in the earlier period. This paper proposes an upgraded approach to upgrading existing services within the Ultra Dense cloud Network (UDN). In this way the cost for the services available there is calculated in advance and a system is developed to manage it effectively. This proposed method clearly shows that the effective cost can be managed by combining high-costservices and low-costservices separately. This will help to manage the various communication processes to achieve high effectiveness in Ultra Dense cloud Network (UDN). The proposed method achieves 93.28% of Cost-effective data Management.49% of Application log failure, 42 % of system log failure, 98% of network log management, 96% of security log management, 99% cost effective management. © 2023 IEEE.

关键词： Cost effectiveness

Optimizing Wireless Resource Management and Synchronization in Digital Twin Networks

学校读者我要写书评

暂无评论

IEEE Internet of Things Journal 2025年

作者： Yu, Hanzhi Liu, Yuchen Yang, Zhaohui Sun, Haijian Chen, Mingzhe University of Miami Department of Electrical and Computer Engineering Frost Institute for Data Science and Computing Coral GablesFL33146 United States North Carolina State University Department of Computer Science RaleighNC27695 United States Zhejiang University College of Information Science and Electronic Engineering Hangzhou310027 China University of Georgia School of Electrical and Computer Engineering AthensGA30602 United States

In this paper, we investigate an accurate synchronization between a physical network and its digital network twin (DNT), which serves as a virtual representation of the physical network. The considered network includes a set of base stations (BSs) that must allocate its limited spectrum resources to serve a set of users while also transmitting its partially observed physical network information to a cloud server to generate the DNT. Since the DNT can predict the physical network status based on its historical status, the BSs may not need to send their physical network information at each time slot, allowing them to conserve spectrum resources to serve the users. However, if the DNT does not receive the physical network information of the BSs over a large time period, the DNT's accuracy in representing the physical network may degrade. To this end, each BS must decide when to send the physical network information to the cloud server to update the DNT, while also determining the spectrum resource allocation policy for both DNT synchronization and serving the users. We formulate this resource allocation task as an optimization problem, aiming to maximize the total data rate of all users while minimizing the asynchronization between the physical network and the DNT. The formulated problem is challenging to solve by traditional optimization methods, as each BS can only observe a partial physical network, making it difficult to find an optimal spectrum allocation strategy for the entire network. To address this problem, we propose a method based on the gated recurrent units (GRUs) and the value decomposition network (VDN). The GRU component allows the DNT to predict future status using the historical data, effectively updating itself when the BSs do not transmit the physical network information. The VDN algorithm enables each BS to learn the relationship between its local observation and the team reward of all BSs, allowing it to collaborate with others in determining whe

关键词： Resource allocation