Approach | Technique | Main Idea |
Wang et al. [43] | FedLDA | Develop a novel framework (named FedLDA) for federated learning of LDA models, as well as a novel LDP mechanism called Random Response with Priori (RRP). |
Girgis et al. [44] | Shuffle method | Develop shuffle method under with only a random permutation of the clients’ responses are received by the server without their association with the clients’ identities. |
Wei et al. [45] | UDP | Develop a novel framework, named UDP, in which each client can achieve adjustable privacy protection levels. |
Zhou et al. [46] | PFLF | Develop a novel framework, named PFLF, in which the client and the central server add noise before sending the data. |
Thapa et al. [47] | SFL | Develop a novel framework SFL that combines federated learning and split learning to achieve high model accuracy and communication efficiency. |
Wu et al. [48] | FedPerGNN | Develop a novel framework FedPerGNN to achieve both effective and privacy-preserving personalization. |
Zhang et al. [50] | PriFedSync | Develop federated f-differential privacy and propose a generic framework for private federated learning. |