Abstract: The significant improvement in patient outcomes that can be achieved with the early identification of chronic kidney disease (CKD). This is because timely therapies can postpone the course of the disease to end-stage renal disease. Using clinical and laboratory characteristics that are routinely obtained, this study examines and analyzes a number of different machine learning algorithms for the early prediction and detection of chronic kidney disease (CKD). In this study, we examine the performance of a basic neural network, logistic regression, support vector machines, and random forest on a benchmark dataset for chronic kidney disease (CKD). Applications include feature selection, handling of missing values, and reduction of class imbalance. Through the utilization of stratified cross-validation, the model that performed the best was able to obtain an accuracy of (placeholder), an F1-score of (placeholder), and AUC-ROC (placeholder). With the help of SHAP values, we additionally offer the interpretability of the model in order to emphasize the clinical characteristics that have the most significant impact. The findings indicate that ensemble tree-based models with proper preprocessing not only offer good performance for early CKD prediction but also aid in clinical adoption through their interpretability.
Keywords: Chronic Kidney Disease, CKD prediction, machine learning, random forest, early prediction.
Title: A Review on Early Prediction and Identification of Chronic Kidney Disease Using Machine Learning Techniques
Author: Jyoti Kiran, Dr. Rohit Goyal, Mr. Rakesh Arya
International Journal of Computer Science and Information Technology Research
ISSN 2348-1196 (print), ISSN 2348-120X (online)
Vol. 14, Issue 4, October 2026 - December 2026
Page No: 1-7
Research Publish Journals
Website: www.researchpublish.com
Published Date: 01-October-2026
