Abstract -Early diagnosis and understanding of Chronic Kidney Disease (CKD) are critical for effective treatment planning, as CKD profoundly impacts kidney function, leading to complications like bone and mineral concerns, low blood pressure, acid-base imbalances, poor nutrition, and neurological disorders. This study explores the application of machine learning (ML) algorithms and various data mining classification methods to predict and diagnose CKD. Leveraging a dataset with 21 features from the UCI Repository, algorithms including Logistic Regression, Decision Tree, SVM, Bagging, Adaboost, Voting Classifier, KNN, Xgboost Gradient Boosting, and Random Forest were employed. Notably, Random Forest exhibited remarkable accuracy at 98.75%. The findings underscore the potential of ML in enhancing CKD identification. This research contributes to the growing body of knowledge on utilizing advanced analytics in healthcare. The 98.75% accuracy achieved by Random Forest emphasizes its efficacy in early CKD detection, offering valuable insights for improved patient care .
Keywords:- Chronic kidney disease, Machine learning, XgBoost classifier, Classification model.
|
DOI:
10.17148/IJARCCE.2024.134160
