Development of Supervised Algorithm for Predicting Chronic Kidney Disease

Chronic Kidney Disease (CKD), also referred to as Chronic Nephritic Sickness, is a serious condition caused by impairment of renal function. Higher levels of fluids and waste accumulate in the body during the later phase of Chronic Kidney Disease and could be life-threatening in many ways. However, the cause and symptoms of this disease could be prevented at the earliest possible time when the necessary medical treatments and measures are undertaken. Hence, certain techniques of machine learning have been introduced to detect Chronic Kidney Failure in its early stages. This research study discusses the important Machine Learning methods sufficient for the detection and diagnosis of CKD to avoid fatal issues. An ensemble method, boosting, has been used that mainly focuses on the concept of the AdaBoost technique. The other two models include K-Nearest Neighbour (KNN) and Random Forest techniques. Notably, Random Forest exhibited the highest accuracy, while KNN and Adaboost demonstrated commendable scores. As a result, healthcare practitioners and policymakers can leverage the insights gained from this research to implement targeted screening programs, design personalized treatment plans, and allocate resources more efficiently to manage and mitigate the impact of CKD on individuals and public health. Ultimately, this paper serves as a valuable tool in advancing the early prediction of CKD and laying the groundwork for proactive healthcare strategies to alleviate the burden of this chronic condition.