<![CDATA[Heart disease is the leading cause of death globally and requires an accurate early prediction system. This study aimed to develop a heart disease classification model using the Support Vector Machine (SVM) method with a Radial Basis Function (RBF) kernel based on the Heart Disease Dataset, which consists of 100 patient records and 13 clinical attributes. The research stages included data preprocessing, feature standardization using StandardScaler, data splitting with an 80:20 ratio, SVM model training, and evaluation using accuracy, precision, recall, F1-score, and confusion matrix metrics. The evaluation results showed that the SVM model with the RBF kernel achieved an accuracy of 85%, with a precision of 0.83 for the negative class and 1.00 for the positive class, recall of 1.00 for the negative class and 0.40 for the positive class, and F1-score of 0.91 for the negative class and 0.57 for the positive class. The confusion matrix produced TN=15, FP=0, FN=3, and TP=2. The low recall of the positive class indicates the model’s limitation in detecting actual heart disease cases (false negatives), mainly caused by class imbalance in the dataset (77 negative : 23 positive) and the limited sample size. Prediction on new clinical data resulted in class 0 (negative), although the data were actually classified as positive, confirming the model’s tendency to produce false negatives in borderline cases. This study highlights the potential of SVM as a tool for early heart disease diagnosis while emphasizing the importance of handling class imbalance and hyperparameter optimization to improve model sensitivity.]]>
