Predicting coronary artery disease: a comparison between two data mining algorithms