Feature Selection Based Data Mining Approach for Coronary Artery Disease Diagnosis

Cardiovascular diseases responsible for many deaths are very common and important health problems. According to World Health Organization, each year 17.7 million people die because of them. Coronary artery disease is the most type cardiovascular that cause serious heart problems in patients, affecting heart’s function negatively. Being aware attributes this will help field-specialist analysis routine laboratory test results a patient coming internal medicine or another unit except cardiology unit. In study, it aimed determine significance coronary by utilizing Stability Selection method. experiments, attributes; ‘Age’, ‘Atypical’, ‘Blood pressure’, ‘Current smoker’, ‘Diastolic murmur’, ‘Dyslipidemia’, ‘Diabetes mellitus’, ‘Ejection fraction’, ‘Erythrocyte sedimentation rate’, ‘Family history’, ‘Hypertension’, ‘Potassium’, ‘Nonanginal’, ‘Pulse ‘Q wave’, ‘Regional wall motion abnormality’, ‘Sex’, ‘St Depression’, ‘Triglyceride’, ‘Tinversion’, ‘Typical chest pain’ ‘Valvular disease’ were found sub-dataset. Besides, performances four traditional machine learning algorithms evaluated detection disease. Logistic Regression algorithm outperformed others with %90.88 value accuracy, 95.18% sensitivity, 81.34% specificity.