Federated Generative Model on Multi-Source Heterogeneous Data in IoT

The study of generative models is a promising branch deep learning techniques, which has been successfully applied to different scenarios, such as Artificial Intelligence and the Internet Things. While in most existing works, are realized centralized structure, raising threats security privacy overburden communication costs. Rare efforts have committed investigating distributed models, especially when training data comes from multiple heterogeneous sources under realistic IoT settings. In this paper, handle challenging problem, we design federated model framework that can learn powerful generator for hierarchical systems. Particularly, our solve problem generation on multi-source two i.e., feature related scenario label scenario. addition, develop synchronous an asynchronous updating methods satisfy application requirements. Extensive experiments simulated dataset real datasets conducted evaluate performance proposed through comparison with state-of-the-arts.