Kidney Disease

Description

The dataset is derived from the publicly available Chronic Kidney Disease dataset originally available on Kaggle (https://www.kaggle.com/datasets/akshayksingh/kidney-disease-dataset). The original dataset contains 400 real-world samples. After preprocessing, 20 features and one target variable indicating whether an individual is classified as a patient or healthy were retained for the present study. Using a calibrated data augmentation procedure (see Supplementary Methods S1), the preprocessed dataset was expanded to 8,314 records while preserving the original class prevalence and feature–target correlation structure. The 8,314 records should therefore not be interpreted as 8,314 independent real-world observations.

Categories

Licence