Evaluation of Symptoms in Heart Disease Patients by using k - Nearest Neighbor Classification

Abstract

In many application domains, classification of complex measurements is essential in a diagnosis process. Correct classification of measurements may in fact be the most critical part of the diagnostic process. In this system, we intend to determine whether a patient has coronary artery disease (CAD) or not and if we have heart disease (CAD) what stage is it by using k - nearest neighbor classification. The k -nearest neighbor (k NN) is a sample and widely used technique which has found in several applications on classification problem. We can get classification accuracy by using k - nearest neighbor algorithm. Experiments were evaluated on some public datasets collected from the Cleveland Clinic Foundation in the UCI (University of California, Irvine) machine learning repository in order to test this system.