Chronic Kidney Disease Prediction Using Hybrid Machine Learning Models

Abstract

The prevalence of chronic kidney disease has steadily increased, becoming a significant issue. Researchers and medical professionals from the Department of Nephrology, the All India Institute of Medical Sciences, and the director general of health services Ministry of Health and Family Welfare, and the Government of India published a report. The estimated prevalence of CKD is 800 per million people. Early detection and characterization are regarded as critical factors in chronic kidney disease management and control. Using effective data mining techniques, physicians can uncover and extract hidden insights from clinical and laboratory patient data, allowing them to accurately identify disease severity stages. This chapter aims to develop a model for predicting risk levels in CKD while taking into account all of the symptoms and causes. Certain solutions can be provided in terms of the dominant characteristics in order to prevent the progression of CKD. Various machine learning approaches can be used to build a model for risk prediction of kidney disease, and their effectiveness can be contrasted in terms of accuracy, specificity, and sensitivity. Before applying any machine learning technique, feature selection must be performed in order to understand the dominant attributes. To select dominant attributes, a feature selection method known as random forest is used.