<![CDATA[Chronic kidney disease (CKD) is a global health issue that impacts quality of life and mortality rates. CKD often shows no symptoms in its early stages, earning it the nickname "silent disease," which complicates early detection efforts. This study aims to develop a classification model for CKD using the K-Nearest Neighbors (KNN) algorithm combined with the Pearson Correlation Coefficient feature selection method to enhance model performance. Feature selection is employed to reduce data dimensionality and prevent overfitting. The Kaggle "Chronic Kidney Disease" dataset is used in this study. Evaluation results show that the model with feature selection achieved an accuracy of 93.37%, precision of 91.9%, recall of 93.37%, and F1-score of 91.48%, while the model without feature selection achieved an accuracy of 91.27%, precision of 87.24%, recall of 91.27%, and F1-score of 88.99%. The contribution of this research is to improve the classification performance of chronic kidney disease by utilizing feature selection methods to achieve a better balance between precision and recall while reducing classification errors.]]>
Copyrights © 2025
