Using Data Mining Techniques for the Diagnosis of Chronic Kidney Disease
Presentation author(s)
Mahima Sharma ’21, Kathmandu, Nepal
Majors: Computer Science; Mathematics
Abstract
Chronic Kidney Disease (CKD) is a growing global health problem that is associated with high healthcare costs and a high mortality rate. Earlier diagnosis is very important since CKD is irreversible in nature. However, its symptoms are difficult to detect until later stages. Data mining techniques can aid physicians in the timely diagnosis of CKD and change the trajectory of the disease’s progression. With this goal in mind, five machine learning algorithms- decision tree, random forest, support vector machine, artificial neural network, and a stacked hybrid model- were used to establish CKD diagnostic models. Then, cluster analyses were carried out to identify common trends in the positively diagnosed samples and to answer whether the clusters conform to the clinical stages of CKD.