Clustering and Classification Approach to Extract the Interdependency among Different Attributes of Cardiac Patients

Abstract

Nowadays we are surrounding with large data related to patient history, test results and reports. Usually, doctors diagnose the disease on the basis of recommended tests. A final recommendation about patient health may involve a lot of factors including patients test results and doctor experience. In this research we have introduced a methodology to extract the interdependency among different attributes of cardiac patients by using the supervised and unsupervised learning. In unsupervised learning we have used the clustering mechanism to extract the interdependency among different attribute of cardiac patients. With the help of useful clustering we can predict the hidden trends in patients. We have used the correlation matrix followed by K-mean (fast) to extract the interesting pattern as well as patient state that will help the practitioner to treat the patient wisely. We have used the different clustering algorithm and analyze the performance of each algorithm in cardiac patient dataset. In supervised learning we have used different classification algorithms to extract the interdependency among different attributes of cardiac patients. We have used ID3, CHAID, Random Tree, Random Forest, Decision Tree and Decision Stump to extract the interdependency among different attributes in cardiac patients. We have performed the comparative analysis of these algorithms; according to `analysis, ID3 gives the best result. This contribution will assist the practitioner to write a wise prescription for cardiac patients.