%0 Journal Article %T Real-data comparison of data mining methods in prediction of coronary artery disease in Iran %J Health Management & Information Science %I Shiraz University of Medical Sciences %Z 2783-302X %A Dekamin, Azam %A Shaibatalhamdi, Ahmad %D 2017 %\ 07/01/2017 %V 4 %N 3 %P 87-94 %! Real-data comparison of data mining methods in prediction of coronary artery disease in Iran %R %X  Introduction: Cardiovascular diseases are currently of broad prevalence and constitute one of the major causes of mortality in different societies. Angiography is one of the most accurate methods to diagnose heart diseases; it incurs high expenses and comes with side effects. Data mining is intended to enable timely prognosis of diseases with the least expenses possible, making use of the patients’ information. The present study aims to provide replies for the question whether it is possible to predict coronary artery diseases with higher efficiency and fewer errors and identify the factors impacting the disease using data mining techniques.Method: In this study, the data under investigation was collected from a number of 303 persons referring to the heart unit in Shahid Rajaie hospital (Iranian hospital) from 2011 to 2013. It included 54 features. Attempts are made to take advantage of a higher number of characteristics which are helpful for diagnosis of diseases. In addition, Information Gain, Gini, and SVM methods were applied to select influential features, and variables with higher weights were chosen for modeling purposes. In the modeling phase, a combination of classification algorithms and ensemble methods was applied to develop a prediction with fewer errors. Rapid Miner Software was adopted to conduct this study.Results: Findings of this research indicated that the suggested model, if weighted by SVM index, had the highest efficiency, i.e. 95.83%. This model, moreover, was able to accurately predict all patients with coronary artery disease in Iran. According to the proposed model and obtained accuracies, weighting with SVM was found to be the most effective filtering method, and age as well as typical and atypical chest pain were identified to be the most effective features of coronary artery disease. (Graph 3)Conclusion: This study can contribute to the diagnosis of influential factors which lead to cardiovascular disease in Iran. Comparison of influential variables showed that chest pain (in its two typical and atypical modes) and patient’s age had the highest weight in this study. It demonstrates that coronary artery disease is more likely to happen in older ages. High blood pressure is also an important factor in outbreak of this disease. That is why measures have to be taken to prevent such occurrence. Diabetes constitutes another influential factor in the outbreak of coronary artery disease to which attention should be paid in primary tests.Keywords: Cardiovascular disease, Coronary Artery Disease, Angiography %U https://jhmi.sums.ac.ir/article_42695_73572d78edd1d56ec6063e0a43ee885e.pdf