• Version
• Download 10
• File Size 468.07 KB
• File Count 1
• Create Date 15 July 2025
• Last Updated 15 July 2025

Chronic Kidney Disease Prediction Using Machine Learning

CH.VASUNDHARA, DALAI MURALI

Assistant Professor, MCA Final Semester, Master of Computer Applications, Sanketika Vidya Parishad Engineering College, Vishakhapatnam, Andhra Pradesh, India

Abstract
This project is designed to predict Chronic Kidney Disease(CKD) using machine learning and provide an easy- to-use web interface for healthcare professionals. It begins by loading a medical dataset and cleaning it by handling missing values and converting incorrect data formats. Next, it performs basic data analysis through visualizations like class distribution, feature correlation, and hemoglobin levels to understand patterns in the data.The project uses a machine learning pipeline where numeric and categorical data are preprocessed separately—missing values are filled, and categorical data is encoded properly. A Random Forest model is then trained to classify whether a patient has CKD or not, using features like age, blood pressure, blood test results, and more. The model's accuracy and performance are evaluated using standard metrics like accuracy score, classification report, and a confusion matrix.After training, the model is saved and integrated into a user-friendly Gradio interface. This allows doctors or users to enter patient details using sliders and dropdown menus and instantly get a prediction about the presence of CKD along with the confidence level. This tool helps in early detection and supports doctors in making faster, data-driven decisions.

Index Terms: Chronic Kidney Disease (CKD) Machine Learning,Random Forest Classifier,Preprocessing,Medical Diagnosis, Feature Engineering,Biomedical Data ,Classification Algorithm, Python,Data Visualization,CKDDetection,Model Evaluation, Early Disease Detection