文章目录论文信息摘要SpreadGNNFramework用于图层次学习的联邦图神经网络图神经网络的联邦多任务学习SpreadGNNDPA-SGD的收敛性质实验总结论文信息SpreadGNN:ServerlessMulti-taskFederatedLearningforGraphNeuralNetworks原文链接：SpreadGNN:ServerlessMulti-taskFederatedLearningforGraphNeuralNetworks：https://arxiv.org/abs/2106.02743摘要GraphNeuralNetworks(GNNs)arethefirstc