Dimension Reduction Effects on Heart Disease Diagnosis Using Machine Learning Techniques

In this method, dimension reduction effects on heart disease prediction have been presented. In addition, four classifiers: K-nearest neighbors (KNN), support vector classifier (SVC), decision tree (DT), random forest (RF) and two types of dimensional reduction models based on principal component analysis (PCA), linear discriminant analysis (LDA) are employed to help in prediction of cardiovascular diseases (SVCs). In Jupyter Notebook, 1025 patients using 14 attributes are utilized that $30 %$ of them are test data. The results of four classifiers are evaluated based on evaluation metrics containing precision, accuracy, recall, F1-score, confusion matrix and receiver operating characteristic (ROC) curve. The results of proposed method demonstrate high performance in heart disease prediction.