mlsanityData to 7 Oct 2026

When Average Calibration Fails: Site-Conditional Federated Conformal Risk Control

arXiv 2606.20115 ↗First posted 18 Jun 2026cs.LG cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Conformal risk control (CRC) provides distribution-free segmentation guarantees by calibrating a prediction-set threshold on held-out data. In federated deployments, the standard approach pools calibration scores into a single threshold. We quantify, on real multi-institutional brain tumor data (FeTS-2022, 1,251 subjects, 20 institutions), a critical failure: naive pooled CRC protects the average hospital but violates coverage at 40% of individual institutions, with the worst site exceeding the target false-negative rate by 7.8 percentage points. We trace this failure to a hidden design choice: the aggregation weights implicitly determine whose coverage is protected. Sample-size weighting optimizes patient-level validity but can sacrifice institution-level reliability; equal-site weighting improves institution-level reliability on this benchmark at comparable efficiency, using only a single scalar per site. We propose risk-curve shrinkage as a principled mechanism: each site transmits its empirical risk curve (G scalars) and a single hyperparameter n0 smoothly interpolates between site-specific local calibration and sample-size-weighted pooled calibration. Leave-one-site-out sensitivity analysis identifies n0=19, achieving 2.7/20 violations at 2.0x stretch. Direct Lagrangian budget optimization fails by concentrating risk on vulnerable hospitals; the finite-sample correction term is essential: removing it triples violations. No patient-level images, masks, or per-volume scores leave any site.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 14 of its 18 references that are in the map. First posted on arXiv 18 Jun 2026.

Builds on

14 of 18 references.

References · 14

  1. 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,592
  2. 2022MONAI: An open-source framework for deep learning in healthcarecited by 438
  3. 2022Conformal Risk Controlcited by 290
  4. 2022Federated Learning Enables Big Data for Rare Cancer Boundary Detectioncited by 83
  5. 2023Federated Conformal Predictors for Distributed Uncertainty Quantificationcited by 46
  6. 2023Conformal Prediction for Federated Uncertainty Quantification Under Label Shiftcited by 29
  7. 2025Conformal Prediction for Image Segmentation Using Morphological Prediction Setscited by 13
  8. 2024Certifiably Byzantine-Robust Federated Conformal Predictioncited by 8
  9. 2025Conformal Risk Control for Semantic Uncertainty Quantification in Computed Tomographycited by 7
  10. 2024FUNAvg: Federated Uncertainty Weighted Averaging for Datasets with Diverse Labelscited by 5
  11. 2025Conditional Conformal Risk Adaptationcited by 5
  12. 2026Efficient Federated Conformal Prediction with Group-Conditional Guaranteecited by 4
2 more
  1. 2026Conformalized Neural Networks for Federated Uncertainty Quantification under Dual Heterogeneitycited by 2
  2. 2024FedStein: Enhancing Multi-Domain Federated Learning Through James-Stein Estimatorcited by 1

Built on by

Not cited yet.