Abstract

CHRONIC KIDNEY DISEASE PREDICTION: A MACHINE LEARNING APPROACH

P. Sai Leela*, Faizan Ansari

Abstract

Chronic Kidney Disease (CKD) is another name for the occurrence of Chronic Renal Disease (CRD). It shows a disease that affects a person's general health and damages the kidneys. Poor illness detection and treatment can lead to end- stage renal disease and the patient's eventual death. Machine Learning (ML) approaches are growing as a useful tool in the medical science sector and are crucial for disease prediction. The proposed project aims to build and validate a predictive model for the outcome of chronic renal illness. This paper extends prior research on chronic kidney disease (CKD) prediction by implementing a scalable and feature-rich machine learning pipeline using a real-world dataset of 20,000 samples. Building upon a foundational study that employed only 400 samples, we introduce additional models, rigorous cross-validation, interaction terms, and improved visual analytics. Our findings suggest that larger data and robust engineering significantly improve prediction performance and generalization capability.

Keywords: chronic kidney disease, classification, Medical Data Analysis, algorithms, random forest classifier, machine learning.