Dynamic Free-Rider Detection in Cross-Silo Federated Learning via Simulated Attack Patterns

Name: Dynamic Free-Rider Detection in Cross-Silo Federated Learning via Simulated Attack Patterns
Uploaded: 2026-10-10T17:29:50.082462+00:00
Duration: 343 s
Description: The S2-WEF method dynamically detects clients who become free-riders during federated learning training without needing pretraining data. It simulates potential attack patterns using previously broadcasted global models to identify suspicious client behavior. This approach offers higher robustness against various free-rider strategies across multiple datasets and attack types.

The S2-WEF method dynamically detects clients who become free-riders during federated learning training without needing pretraining data. It simulates potential attack patterns using previously broadcasted global models to identify suspicious client behavior. This approach offers higher robustness against various free-rider strategies across multiple datasets and attack types.

Open the full episode