Studies on Application of Support Vector Machine in Diagnose of Coronary Heart Disease

Right now, the diagnosis of coronary heart disease is mostly from experienced physician´s judgment. How to use computer intelligent algorithms to aid in diagnosing of coronary heart disease has been a hot research of machine learning. This article will apply support vector machine (SVM) method which is based on the statistical learning theory to the diagnosis of coronary heart disease. On the basis of original data pre-processing and feature extraction, classifiers with different kernel are selected to classify the test data, followed by a comparison of classification results which show that the accuracy of classifiers with radial basis function is the highest. To select the best parameters of kernel function and penalty factor, Grid Search Method of optimizing parameters is used, which makes the classifier achieve the highest classification accuracy.