Abstract
Chronic Kidney Disease (CKD) has become a major health concern worldwide, affecting
millions of people and often leading to fatal outcomes. Early detection and accurate prediction of CKD
are crucial for timely intervention and effective management. In this study, we propose a machine
learning-based prediction system for CKD, leveraging clinical and demographic data to predict the risk
of CKD onset. The system utilizes a dataset comprising comprehensive clinical features, including age,
gender, blood pressure, serum creatinine, and other relevant parameters.
We explore various machine learning algorithms, such as Random Forest, Support Vector Machine
(SVM), Decision Tree, K-Nearest Neighbors (KNN), and Logistic Regression, to develop and compare
the performance of our prediction model. The dataset was preprocessed, including handling missing
values, normalization, and feature selection, to ensure optimal model performance. Evaluation metrics,
including accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve
(AUC-ROC), were employed to assess the models' performance.
Our results indicate that the Random Forest algorithm outperforms other models with an accuracy of over
95% and an AUC-ROC of 0.98. The developed prediction system demonstrates promising results in
identifying individuals at risk of developing CKD. The proposed system could be integrated into clinical
practice to assist healthcare professionals in making informed decisions, leading to early detection and
timely intervention, thereby improving patient outcomes and reducing the burden on healthcare systems.