Federated Learning Security: Understanding Malicious Clients

Clarifying the confusion around poisoning attacks in Federated Learning by presenting a spectrum of adversary models and proposing a hybrid model for more effective attacks.