2.2 Federated Multi-Task Learning with Graph Neural Networks 基于多任务学习(MTL),定义 centralized federated graph MTL(FedGMTL)如下: \mathop{\min}_{\theta,\Psi,\Phi_{pool},\Phi_{task}}\sum_{k=1}^K\frac{1}{N_k}\sum_{i=1}^{N_k}\mathcal{L}(\hat{\mathbf{y}}_i^{(k)},\ma...
Ditto是通过减少对全局模型的依赖来进行联邦个性化的。除了Ditto还有其他个性化方法也能达到相似的技术,最后实验部分也有提到,那为啥深入研究Ditto呢?因为效果好,而且简单 Ditto与将朝其平均值正则化个性化模型的工作有很大的关系,类似于经典的均值正则化MTL; 不同之处在于Ditto是正则化全局模型,而不是平均的个性化模型。
因此，联邦学习与这种联合学习是完全不同的概念。还有一种学习方法叫做“多任务学习”(multitask learning)[8,9]，它是迁移学习的一个子方向，旨在有多个学习目标并部分共用数据的情况下，尽量多地利用共有模型部分来提高学习效果。多任务学习对数据安全和隐私也没有提出要求，而是一种机器学习算法。联邦学习概述 什么...
Federated learning (FL) for medical image segmentation becomes more challenging in multi-task settings where clients might have different categories of labels represented in their data. For example, one client might have patient data with "healthy'' pancreases only while datasets from other clients ...
Similarly, Yu et al. (2020) build a federated multi- task learning framework for smart home IOT to automatically learn users’ behavior patterns, which could effectively detect physical hazards. Furthermore, Liu, Wang, Liu, and Xu (2020) proposed a data fusion approach based on FL for robots...
Federated multi-task learning (FMTL) has emerged as a natural choice to capture the statistical diversity among the clients in federated learning. To unleash the potential of FMTL beyond statistical diversity, we formulate a new FMTL problem FedU using Laplacian regularization, which can explicitly...
a novel multi-task federated training framework capable of operating in the presence of partial labels and absence of a central server for the first time in the literature. SpreadGNN extends federated multi-task learning to realistic serverless settings for GNNs, and utilizes a novel optimization al...
当以“federated personal”为关键词,在 dblp 上进行文献检索的时候,会发现个性化联邦学习(PFL, Personalized Federated Learning)的研究在2021年的有明显增加。 经小鱼的查询,至22年4月16日针对个性化联邦学习的综述文章仅两篇[1][2],其中...
Repository files navigation README Apache-2.0 license mocha Federated Multi-Task LearningAbout Federated Multi-Task Learning Resources Readme License Apache-2.0 license Activity Stars 129 stars Watchers 5 watching Forks 39 forks Report repository Releases No releases published Packages No pack...
现有联邦学习个性化技术《2020-IEEE-Survey of Personalization Techniques for Federated Learning》包括多任务学习《2017-NIPSFederated Multi-Task Learning》,基础层+个性化层《2019-Federated learning with personalization layers》,全局和局部模型的混合《2020-Federated learning of a mixture of global and local ...