Decentralised federated learning with adaptive partial gradient aggregation

Jiang, Jingyan; Hu, Liang

doi:10.1049/trit.2020.0082

Cited by 52 publications

(17 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…e algorithm will locate the points over which the gray level is not changing, and it works on all the planes of the 3D image. For the scanning of the 3D image, 3 × 3 × 3 size of the sliding window [42,43]…”

Section: Wireframe Of the Imagementioning

confidence: 99%

Efficient 3D AlexNet Architecture for Object Recognition Using Syntactic Patterns from Medical Images

Rani

Ghai

Kumar

et al. 2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

In computer vision and medical image processing, object recognition is the primary concern today. Humans require only a few milliseconds for object recognition and visual stimulation. This led to the development of a computer-specific pattern recognition method in this study for identifying objects in medical images such as brain tumors. Initially, an adaptive median filter is used to remove the noise from MRI images. Thereafter, the contrast image enhancement technique is used to improve the quality of the image. To evaluate the wireframe model, the cellular logic array processing (CLAP)-based algorithm is then applied to images. The basic patterns of three-dimensional (3D) images are then identified from the input image by scanning the whole image. The frequency of these patterns is also used for object classification. A deep neural network is then utilized for the classification of brain tumor. In the proposed model, the syntactic pattern recognition technique is used to find the feature vector and 3D AlexNet is used for brain tumor classification. To evaluate the performance of the proposed work, three benchmark brain tumor datasets are used, i.e., Figshare, Brain MRI Kaggle, and Medical MRI datasets and BraTS 2019 dataset. The comparative analyses reveal that the proposed brain tumor classification model achieves significantly better performance than the existing models.

show abstract

Section: Wireframe Of the Imagementioning

confidence: 99%

Efficient 3D AlexNet Architecture for Object Recognition Using Syntactic Patterns from Medical Images

Rani

Ghai

Kumar

et al. 2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

show abstract

“…However, without a central server, the holistic perspective is impractical, it is hard for clients to form such a topology with limited observations. Second, communication of P2P FL can be more efficient by sparsification [26], adaptive partial gradient aggregation [27], and using max-plus linear system theory to compute throughput [28].…”

Section: Related Workmentioning

confidence: 99%

Mining Latent Relationships among Clients: Peer-to-peer Federated Learning with Adaptive Neighbor Matching

Li¹,

Lu²,

Luo³

et al. 2022

Preprint

View full text Add to dashboard Cite

In federated learning (FL), clients may have diverse objectives, merging all clients' knowledge into one global model will cause negative transfers to local performance. Thus, clustered FL is proposed to group similar clients into clusters and maintain several global models. Nevertheless, current clustered FL algorithms require the assumption of the number of clusters, they are not effective enough to explore the latent relationships among clients. However, we take advantage of peer-to-peer (P2P) FL, where clients communicate with neighbors without a central server and propose an algorithm that enables clients to form an effective communication topology in a decentralized manner without assuming the number of clusters. Additionally, the P2P setting will release the concerns caused by the central server in centralized FL, such as reliability and communication bandwidth problems. In our method, 1) we present two novel metrics for measuring client similarity, applicable under P2P protocols; 2) we devise a two-stage algorithm, in the first stage, an efficient method to enable clients to match same-cluster neighbors with high confidence is proposed; 3) then in the second stage, a heuristic method based on Expectation Maximization under the Gaussian Mixture Model assumption of similarities is used for clients to discover more neighbors with similar objectives. We make a theoretical analysis of how our work is superior to the P2P FL counterpart and extensive experiments show that our method outperforms all P2P FL baselines and has comparable or even superior performance to centralized cluster FL. Moreover, results show that our method is much effective in mining latent cluster relationships under various heterogeneity without assuming the number of clusters and it is effective even under low communication budgets.

show abstract

“…• FedPGA [15] is a decentralised aggregation algorithm developed from FedAvg. The devices in FedPGA exchange partial gradients rather than full model weights.…”

Section: Related Patterns: Model Co-versioning Registry Incentive Reg...mentioning

confidence: 99%

Architectural Patterns for the Design of Federated Learning Systems

Lo¹,

Lu²,

Zhu³

et al. 2021

Preprint

View full text Add to dashboard Cite

Federated learning has received fast-growing interests from academia and industry to tackle the challenges of data hungriness and privacy in machine learning. A federated learning system can be viewed as a large-scale distributed system with different components and stakeholders as numerous client devices participate in federated learning. Designing a federated learning system requires software system design thinking apart from machine learning knowledge. Although much effort has been put into federated learning from the machine learning technique aspects, the software architecture design concerns in building federated learning systems have been largely ignored. Therefore, in this paper, we present a collection of architectural patterns to deal with the design challenges of federated learning systems. Architectural patterns present reusable solutions to a commonly occurring problem within a given context during software architecture design. The presented patterns are based on the results of a systematic literature review and include three client management patterns, four model management patterns, three model training patterns, and four model aggregation patterns. The patterns are associated to particular state transitions in a federated learning model lifecycle, serving as a guidance for effective use of the patterns in the design of federated learning systems.

show abstract

Decentralised federated learning with adaptive partial gradient aggregation

Cited by 52 publications

References 17 publications

Efficient 3D AlexNet Architecture for Object Recognition Using Syntactic Patterns from Medical Images

Efficient 3D AlexNet Architecture for Object Recognition Using Syntactic Patterns from Medical Images

Mining Latent Relationships among Clients: Peer-to-peer Federated Learning with Adaptive Neighbor Matching

Architectural Patterns for the Design of Federated Learning Systems

Contact Info

Product

Resources

About