Our work is motivated by the lack of a comprehensive survey that organizes attacks based on the specific targets of the attacks within the federated learning ecosystem. Compared to existing surveys on FL, our work introduces a novel taxonomy of attacks that is specifically based on the attack tar...
However, recent research has shown that the training process in federated learning is also susceptible to various degrees of attacks, leading to data leakage. Therefore, researchers in the field have proposed various defense mechanisms for federated learning. In this paper, we introduce different ...
A Survey on Federated Learning: The Journey From Centralized to Distributed On-Site Learning and Beyond Authors Sawsan AbdulRahman, Hanine Tout, Hakim
4.3 Robustness to Targeted Attacks 锁定目标攻击比如植入后门,然后使用trigger启动后门,此时模型性能不会下降,数据也正常,因此这种攻击非常隐蔽,讨论最多的是后门攻击,可以通过女巫攻击增强 后门攻击:即在模型中植入一些后门,模型表现完全正常,但trigger启动时就会导致误检,现有的就是模型中毒,能植入一些语义后门导致误检...
A survey on federated learning Authors Chen Zhang, Yu Xie, Hang Bai, Bin Yu, Weihong Li, Yuan Gao Keywords Federated learning; Privacy protection; Mac
A survey on privacy-preserving federated learning against poisoning attacks Federated learning (FL) is designed to protect privacy of participants by not allowing direct access to the participants' local datasets and training proce... F Xia,W Cheng - 《Cluster Computing》 被引量: 0发表: 2024年 ...
Threats, Attacks, and Defenses in Machine Unlearning: A Survey arXiv 2024 Machine Unlearning: Solutions and Challenges TETCI 2024 A Survey on Federated Unlearning: Challenges, Methods, and Future Directions arXiv 2023 A Survey of Federated Unlearning: A Taxonomy, Challenges and Future Directions arXi...
Federated transfer learning:样本空间和特征空间均不相同,有人用秘密分析技术提高通信效率,应用比如不同疾病治疗方式可迁移; 3. Evolution of FL 现在主要两条研究方向:提升效率和精度的算法优化,保护数据安全的隐私优化; 算法优化:通信负担,数据异质,结构异质 ...
Federated learning is known to be vulnerable to both security and privacy issues. Existing research has focused either on preventing poisoning attacks from... T Nguyen,MT Thai - IEEE/ACM Transactions on Networking: A Joint Publication of the IEEE Communications Soceity, the IEEE Computer Society, ...
To ensure that we out-turn the benefits of FL over risks and utilize the features of FL properly, we have an immediate need to be on top of this area of research to investigate all possible security and privacy attacks on FL. Without precise information and clear vision, FL may be ...