Different Corruptions, Different Signals: Uncertainty and Loss in Federated Data Quality

The paper investigates how two signals—input‑conditional uncertainty and prediction‑label loss—detect different types of data corruption in federated learning. Experiments on ResNet‑20 with CIFAR‑10 and SVHN show that prediction‑label loss excels at spotting persistent random label flips, while expected‑entropy uncertainty better identifies additive image noise. The authors argue that effective federated data‑quality assessment must match the chosen signal to the specific corruption type rather than rely solely on uncertainty measures.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

• data-engineering