Application of K-Means Technique in Data Mining to Cluster Hemodialysis Patients

Abstract

Hemodialysis patients with end stage renal disease are often hospitalized due to infection, cardiovascular syndromes, or cancer. Through interviews with the doctors at a well-known hospital for kidney disease, a list of important variables including the clinical lab test results, the demographic and the socio-economic information was developed. After finalizing the list with the doctors, the needed data of 50 patients was provided anonymously so that real data can be used for this research. The review of literature reveals that no other research considered all of these measures used in this work simultaneously. In the next stage, the data was processed and prepared to be used as the input data for the implementation of data mining techniques. The clustering of patients via the K-Means technique is explained here. The results obtained from the data mining model in this research can be useful for the decision makers and doctors. Implementing the model seems promising and analyzing the results proved this point and delivered interesting grouping of patients. Also, it can be claimed that this research is a good prologue for future researches.