Prognostic models show poorer external calibration with larger distributional shifts between cohorts, but meta-analysis weighting and similarity-based selection improve transportability and clinical utility.
specialization under concept shift
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Robustifying and Selecting Cohort-Appropriate Prognostic Models under Distributional Shifts
Prognostic models show poorer external calibration with larger distributional shifts between cohorts, but meta-analysis weighting and similarity-based selection improve transportability and clinical utility.