Predicting Chronic Kidney Disease Progression Using Classification and Ensemble Learning

A serious medical problem that impacts millions of people globally is chronic kidney disease, or CKIDD. Accurate CKIDD progression prediction is essential for disease management and early intervention. Machine learning (ML) has shown great potential in predicting the course of CKIDD by employing a variety of medical, laboratory, and demographic factors. In our research, we suggested an ensemble approach for the increased level of accuracy. We contrast the outcome with well-known machine learning models such as neural networks, Dec-Trees, random forest, and support vector machines. On a sizable dataset of CKIDD patients, where good progression outcomes are identified, our model is trained. Our model is trained on a big dataset of CKIDD patients and proposed to find the result that the patient is affected or not. The chronic kidney disease dataset from Kaggle is utilized to test the model. Standard metrics including accuracy, precision, recall, F1-score, and area under receiver operating characteristic curve (AUC-ROC) are considered for proving the capability of the proposed work. In comparison to the conventional method, our model's experiment results demonstrate 100% accuracy in pattern prediction.