Tag
The paper presents a two-part study using machine learning on large-scale telehealth data to classify self-reported chronic kidney disease status and identify key psychosocial and medical predictors, achieving balanced accuracy around 72-76% with SHAP analysis for interpretability.