Answers are grounded in this paper and cite their sources.
Correcting for class imbalance is not always necessary and may even be harmful to clinical prediction models which aim to produce reliable risk estimates on an individual basis.
No takes yet. Share an insight, caveat, or question.
Carriero et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: