CKD prediction system implementing MICE imputation, Borderline-SMOTE, RFE/Boruta feature selection, and ensemble learning models including Random Forest, XGBoost, and LightGBM.
-
Updated
Jun 18, 2026 - Jupyter Notebook
CKD prediction system implementing MICE imputation, Borderline-SMOTE, RFE/Boruta feature selection, and ensemble learning models including Random Forest, XGBoost, and LightGBM.
Pipeline di ML clinico per il pre-screening opportunistico della malattia renale cronica. Senza esami renali (zero creatinina o urine), sfrutta i parametri ematochimici di routine per identificare soggetti a rischio KDIGO, riducendo oltre il 43% dei test inutili e guidando il triage di primo livello prima del laboratorio specialistico.
To associate your repository with the ckd-detection topic, visit your repo's landing page and select "manage topics."