Filters








8,782 Hits in 4.4 sec

Dynamic Gradient Aggregation for Federated Domain Adaptation [article]

Dimitrios Dimitriadis, Kenichi Kumatani, Robert Gmyr, Yashesh Gaur, Sefik Emre Eskimez
2021 arXiv   pre-print
In this paper, a new learning algorithm for Federated Learning (FL) is introduced.  ...  The proposed scheme is based on a weighted gradient aggregation using two-step optimization to offer a flexible training pipeline.  ...  Dynamic Gradient Aggregation: A novel algorithm for de-emphasizing batches of "bad" data is presented.  ... 
arXiv:2106.07578v1 fatcat:xgbkwqfxsvas5epyzvrm2q7jrq

Aggregating Gradients in Encoded Domain for Federated Learning [article]

Dun Zeng, Shiyu Liu, Siqi Liang, Zonghang Li, Zenglin Xu
2022 arXiv   pre-print
To mitigate the cost, we propose the framework, which enables the server to aggregate gradients in an encoded domain without accessing raw gradients of any single client.  ...  Malicious attackers and an honest-but-curious server can steal private client data from uploaded gradients in federated learning.  ...  Encoded Domain Aggregation This part is to demonstrate our strategies for building an error-bounded and adaptive encoder-decoder network (E c , D c ) for encoded domain aggregation.  ... 
arXiv:2205.13216v2 fatcat:yxxquhm5z5g2zlsqade6knxrmi

Federated Adversarial Domain Adaptation [article]

Xingchao Peng, Zijun Huang, Yizhe Zhu, Kate Saenko
2019 arXiv   pre-print
Empirically, we perform extensive experiments on several image and text classification tasks and show promising results under unsupervised federated domain adaptation setting.  ...  In this work, we present a principled approach to the problem of federated domain adaptation, which aims to align the representations learned among the different nodes with the data distribution of the  ...  (Weighted error bound for federated domain adaptation).  ... 
arXiv:1911.02054v2 fatcat:obqtydte5nczner5glyzsunleu

IOP-FL: Inside-Outside Personalization for Federated Medical Image Segmentation [article]

Meirui Jiang, Hongzheng Yang, Chen Cheng, Qi Dou
2022 arXiv   pre-print
Our inside personalization is achieved by a lightweight gradient-based approach that exploits the local adapted model for each client, by accumulating both the global gradients for common knowledge and  ...  local gradients for client-specific optimization.  ...  Please note that the aggregation is naturally performed in the federated learning after certain local training epochs, the calculation of the local adapted model based on global and local gradients does  ... 
arXiv:2204.08467v1 fatcat:j7k6cox7hnadzayhi7xh44576m

Federated Adaptation of Reservoirs via Intrinsic Plasticity [article]

Valerio De Caro, Claudio Gallicchio, Davide Bacciu
2022 arXiv   pre-print
The former is a gradient-based method for adapting the reservoir's non-linearity in a local and unsupervised manner, while the latter provides the framework for learning in the federated scenario.  ...  In particular, our proposal focuses on the adaptation of reservoirs by combining Intrinsic Plasticity with Federated Averaging.  ...  Our proposal is based on Intrinsic Plasticity [6] , an existing algorithm for adapting the dynamics of a reservoir with respect to the input sequence, and Federated Averaging, a client-server algorithm  ... 
arXiv:2206.11087v1 fatcat:mtahypskhze47cvv6d3zfac5w4

Federated Transfer Learning: concept and applications [article]

Sudipan Saha, Tahir Ahmad
2021 arXiv   pre-print
Among different variants of the federated learning, noteworthy is federated transfer learning (FTL) that allows knowledge to be transferred across domains that do not have many overlapping features and  ...  Federated learning has emerged as a possible solution to this problem in the last few years without compromising user privacy.  ...  [52] presented a dynamic gradient aggregation (DGA) method which weights the local gradients during aggregation step.  ... 
arXiv:2010.15561v3 fatcat:3udixrhta5btlb7w7r4fomwpzu

FedMed: A Federated Learning Framework for Language Modeling

Xing Wu, Zhaowang Liang, Jianjia Wang
2020 Sensors  
We propose a novel Federated Mediation (FedMed) framework with the adaptive aggregation, mediation incentive scheme, and topK strategy to address the model aggregation and communication costs.  ...  However, there are two major problems in the federated optimization for the prediction: (1) aggregating model parameters on the server-side and (2) reducing communication costs caused by model weights  ...  Acknowledgments: We appreciate the High Performance Computing Center of Shanghai University, and Shanghai Engineering Research Center of Intelligent Computing System (No. 19DZ2252600) for providing the  ... 
doi:10.3390/s20144048 pmid:32708152 fatcat:yibkcghyknbanju3ib2cfeavlm

Addressing Client Drift in Federated Continual Learning with Adaptive Optimization [article]

Yeshwanth Venkatesha, Youngeun Kim, Hyoungseob Park, Yuhang Li, Priyadarshini Panda
2022 arXiv   pre-print
However, there is little attention towards additional challenges emerging when federated aggregation is performed in a continual learning system.  ...  Federated learning has been extensively studied and is the prevalent method for privacy-preserving distributed learning in edge devices.  ...  The federated aggregation can be further generalized to more adaptive gradient descent algorithms. This family of algorithms is described as FedOpt in [38] .  ... 
arXiv:2203.13321v1 fatcat:tnx2zsq2cnfmrh3nkl3heye3nu

Federated Graph Machine Learning: A Survey of Concepts, Techniques, and Applications [article]

Xingbo Fu, Binchi Zhang, Yushun Dong, Chen Chen, Jundong Li
2022 arXiv   pre-print
Federated Graph Machine Learning (FGML) is a promising solution to tackle this challenge by training graph machine learning models in a federated manner.  ...  In addition, we summarize the real-world applications of FGML from different domains and introduce open graph datasets and platforms adopted in FGML.  ...  ., ResNet [42] and DenseNet [45] ), domain and class-specific features and the GNN models are also transmitted for aggregation during federated optimization.  ... 
arXiv:2207.11812v1 fatcat:kexw3btwjngijd3tobbi7wqgfy

Artificial intelligence across company borders [article]

Olga Fink, Torbjørn Netland, Stefan Feuerriegel
2021 arXiv   pre-print
Combining federated learning with domain adaptation can provide a solution to this problem by enabling effective cross-company AI without data disclosure.  ...  At the same time, a substantial potential for utilizing AI across company borders has remained largely untapped.  ...  For instance, if there is no significant domain shift between the applications of two companies, federated learning can be applied without combining it with domain adaptation.  ... 
arXiv:2107.03912v1 fatcat:vfk7mmbbnrc6nbing63vyy7uau

Federated Progressive Sparsification (Purge, Merge, Tune)+ [article]

Dimitris Stripelis, Umang Gupta, Greg Ver Steeg, Jose Luis Ambite
2022 arXiv   pre-print
To improve federated training of neural networks, we develop FedSparsify, a sparsification strategy based on progressive weight magnitude pruning. Our method has several benefits.  ...  "pseudo-gradients" upon which the server applies different adaptive optimization techniques.  ...  For dynamic pruning, we compare against PruneFL [Jiang et al., 2020] that uses an adaptive model pruning technique with the initial pruning stage performed on a randomly selected client from the federation  ... 
arXiv:2204.12430v1 fatcat:4x5r7atimvdonalw6ztv7uzuru

Federated Dynamic GNN with Secure Aggregation [article]

Meng Jiang and Taeho Jung and Ryan Karl and Tong Zhao
2020 arXiv   pre-print
In this work, we introduce Federated Dynamic Graph Neural Network (Feddy), a distributed and secured framework to learn the object representations from multi-user graph sequences: i) It aggregates structural  ...  Given video data from multiple personal devices or street cameras, can we exploit the structural and dynamic information to learn dynamic representation of objects for applications such as distributed  ...  Here α is the weight for aggregating spatial information from neighbors; β is the weight for aggregating dynamic information from neighboring graphs. The aggregation is applied every 10 epochs.  ... 
arXiv:2009.07351v1 fatcat:7j6bklorfbegdddgqj42lsknwe

Memory-aware curriculum federated learning for breast cancer classification [article]

Amelia Jiménez-Sánchez, Mickael Tardy, Miguel A. González Ballester, Diana Mateus, Gemma Piella
2021 arXiv   pre-print
Our approach is combined with unsupervised domain adaptation to deal with domain shift while preserving data privacy. We evaluate our method with three clinical datasets from different vendors.  ...  In this work, we define a memory-aware curriculum learning method for the federated setting.  ...  Regarding the local model aggregation, one could deploy a CL-based adaptive weighting for clients based on a dynamic scoring function taking into account meta-information [21] , and in this way, help  ... 
arXiv:2107.02504v1 fatcat:jxtases4ojb7jd652r2g5fwoju

Auto-FedRL: Federated Hyperparameter Optimization for Multi-institutional Medical Image Segmentation [article]

Pengfei Guo, Dong Yang, Ali Hatamizadeh, An Xu, Ziyue Xu, Wenqi Li, Can Zhao, Daguang Xu, Stephanie Harmon, Evrim Turkbey, Baris Turkbey, Bradford Wood (+5 others)
2022 arXiv   pre-print
In this work, we propose an efficient reinforcement learning~(RL)-based federated hyperparameter optimization algorithm, termed Auto-FedRL, in which an online RL agent can dynamically adjust hyperparameters  ...  Federated learning (FL) is a distributed machine learning technique that enables collaborative model training while avoiding explicit data sharing.  ...  This general FL optimization formulation refers to adaptive federated optimization [38] .  ... 
arXiv:2203.06338v1 fatcat:ig3amhtkkzbq3jnk5m6pnowj3e

Voting-based Approaches For Differentially Private Federated Learning [article]

Yuqing Zhu, Xiang Yu, Yi-Hsuan Tsai, Francesco Pittaluga, Masoud Faraki, Manmohan chandraker, Yu-Xiang Wang
2021 arXiv   pre-print
Differentially Private Federated Learning (DPFL) is an emerging field with many applications.  ...  , instead of averaging the gradients, which avoids the dimension dependence and significantly reduces the communication cost.  ...  Federated learning with domain adaptation has been studied in Peng et al. (2019b) , where they propose a dynamic attention model to adjust the contribution from each source (agent) collaboratively.  ... 
arXiv:2010.04851v2 fatcat:ff5qqlgdonhefexhjoc5fep32q
« Previous Showing results 1 — 15 out of 8,782 results