PRISM-FCP: Byzantine-Resilient Federated Conformal Prediction via Partial Sharing
2026-02-20 • Machine Learning
Machine Learning
AI summaryⓘ
The authors introduce PRISM-FCP, a new method to make federated learning more reliable even when some participants try to sabotage the process (called Byzantine attacks). Unlike previous methods that only protect the final calibration step, their approach defends during both training and calibration by sharing only part of the model updates and by filtering suspicious calibration data. This helps keep prediction results accurate and reliable without needing extra communication. Experiments show that their method maintains trustworthy predictions and tighter confidence intervals under attacks compared to existing solutions.
Federated LearningByzantine AttacksConformal PredictionRobust CalibrationPartial Model SharingNonconformity ScoresMaliciousness DetectionCommunication EfficiencyMean-Square ErrorPrediction Intervals
Authors
Ehsan Lari, Reza Arablouei, Stefan Werner
Abstract
We propose PRISM-FCP (Partial shaRing and robust calIbration with Statistical Margins for Federated Conformal Prediction), a Byzantine-resilient federated conformal prediction framework that utilizes partial model sharing to improve robustness against Byzantine attacks during both model training and conformal calibration. Existing approaches address adversarial behavior only in the calibration stage, leaving the learned model susceptible to poisoned updates. In contrast, PRISM-FCP mitigates attacks end-to-end. During training, clients partially share updates by transmitting only $M$ of $D$ parameters per round. This attenuates the expected energy of an adversary's perturbation in the aggregated update by a factor of $M/D$, yielding lower mean-square error (MSE) and tighter prediction intervals. During calibration, clients convert nonconformity scores into characterization vectors, compute distance-based maliciousness scores, and downweight or filter suspected Byzantine contributions before estimating the conformal quantile. Extensive experiments on both synthetic data and the UCI Superconductivity dataset demonstrate that PRISM-FCP maintains nominal coverage guarantees under Byzantine attacks while avoiding the interval inflation observed in standard FCP with reduced communication, providing a robust and communication-efficient approach to federated uncertainty quantification.