Heart-Disease Diagnosis via Support Vector Machine-Based Approaches

Heart-disease diagnosis is widely studied by researchers all over the world, since it is the primary cause of deaths. There exist many challenges in heart-disease diagnosis, such as huge amount of data, high data dimension, large noise interference, etc, which point to the suitability of using data-driven approaches. This paper presents two dimension-reduction methodologies based on support vector machine (SVM), to diagnose heart disease. The most relevant features for diagnosis are achieved by support vector machine-recursive feature elimination (SVM-RFE) method. Then principal component analysis-support machine (PCA-SVM) is also used for heart-disease diagnosis. The best classification accuracy 88.24% is obtained by PCA-SVM via Radial Basis Function (RBF) kernel using only 6 principal components.