mlsanityData to 14 Oct 2026

CHOIR: heterogeneity-aware conformal prediction for crash injury severity across driver safety strata

arXiv 2609.11592 ↗First posted 10 Sep 2026stat.ML cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Transportation agencies increasingly predict crash-injury severity with statistical and machine-learning models, but these models do not state how often their output contains the recorded injury level or for which groups of drivers it fails, a gap that matters most for motorcyclists and unrestrained drivers. This study develops and evaluates a certification layer that gives any fitted severity model a finite-sample, distribution-free coverage guarantee within prespecified safety strata. The layer, CHOIR (Conformal Heterogeneity-aware Ordinal Inference with Risk control), combines groupwise and weighted conformal prediction with conformal risk control to return contiguous KABCO intervals, and adds a declared sensitivity analysis for medically assessed injury and bounds on fatal omission. It is evaluated on 4.04 million Texas crashes from 2017-2023, one sampled driver per crash, with seven base models from the ordered logit to a tabular foundation model, and on held-out counties and later years. Under one pooled threshold every model reaches 0.90 coverage overall but covers motorcyclists or unrestrained drivers at 0.868 or lower, and class-balanced gradient boosting covers unrestrained drivers at only 0.374. Calibration within four safety strata places all 28 model-by-stratum estimates between 0.898 and 0.907, at the cost of sets spanning 3.6 to 4.6 of five categories for these groups, and the certified ordered logit is within 0.03 categories of the narrowest model. Injury-model coverage should therefore be certified within safety groups rather than on average, calibration rather than model complexity determines validity, and a statewide threshold should not be applied to small rural counties without local calibration data.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 24 of its 53 references that are in the map. First posted on arXiv 10 Sep 2026.

Builds on

24 of 53 references.

References · 24

  1. 2016Distribution-Free Predictive Inference For Regressioncited by 668
  2. 2019Conformal Prediction Under Covariate Shiftcited by 519
  3. 2022Conformal prediction beyond exchangeabilitycited by 355
  4. 2016Least Ambiguous Set-Valued Classifiers with Bounded Error Levelscited by 272
  5. 2021Testing for Outliers with Conformal p-valuescited by 134
  6. 2022Improving Trustworthiness of AI Disease Severity Rating in Medical Imaging with Ordinal Conformal Prediction Setscited by 38
  7. 2018Distribution-Free Prediction Sets for Two-Layer Hierarchical Modelscited by 35
  8. 2023Conformal prediction under ambiguous ground truthcited by 31
  9. 2023Adaptive conformal classification with noisy labelscited by 25
  10. 2024Group-Weighted Conformal Predictioncited by 14
  11. 2022Predictive Inference with Weak Supervisioncited by 13
  12. 2025Conformal Inference under High-Dimensional Covariate Shifts via Likelihood-Ratio Regularizationcited by 12
12 more
  1. 2024Conformal Risk Control for Ordinal Classificationcited by 8
  2. 2025Conformal Prediction of Classifiers with Many Classes based on Noisy Labelscited by 5
  3. 2025Exploring the Noise Robustness of Online Conformal Predictioncited by 5
  4. 2025Non-exchangeable Conformal Prediction with Optimal Transport: Tackling Distribution Shifts with Unlabeled Datacited by 5
  5. 2026Weight Clipping for Robust Conformal Inference under Unbounded Covariate Shiftscited by 4
  6. 2026Conformal Risk Control for Safety-Critical Wildfire Evacuation Mapping: A Comparative Study of Tabular, Spatial, and Graph-Based Modelscited by 3
  7. 2024Distribution-free Conformal Prediction for Ordinal Classificationcited by 2
  8. 2025Prediction Sets and Conformal Inference with Interval Outcomescited by 2
  9. 2025Provably Minimum-Length Conformal Prediction Sets for Ordinal Classificationcited by 1
  10. 2026Socio-Conformal Calibration in Complex Survey Data: Marginal Validity Is Not Enough for Subgroup Reliabilitycited by 1
  11. 2026Reliable Conformal Prediction for Ordinal Classification Using the Ranked Probability Scorecited by 1
  12. 2026A Quiet Failure in Calibrated Virtual Screening: Marginal Conformal Prediction Under-Covers the Minority Class, and a Class-Conditional Fix Recovers Itcited by 1

Built on by

Not cited yet.