Machine Learning Application to Predict the Risk of Coronary Artery Atherosclerosis

Coronary artery disease is the leading cause of death in the world. In this research, we propose an algorithm based on the machine learning techniques to predict the risk of coronary artery atherosclerosis. A ridge expectation maximization imputation (REMI) technique is proposed to estimate the missing values in the atherosclerosis databases. A conditional likelihood maximization method is used to remove irrelevant attributes and reduce the size of feature space and thus improve the speed of the learning. The STULONG and UCI databases are used to evaluate the proposed algorithm. The performance of heart disease prediction for two classification models is analyzed and compared to previous work. Experimental results show the improved accuracy percentage of risk prediction of our proposed method compared to other works. The effect of missing value imputation on the prediction performance is also evaluated and the proposed REMI approach performs significantly better than conventional techniques.