Enhancing the Diagnosis of Cardiovascular Disease: A Comparative Examination of Support Vector Machine and Artificial Neural Network Models Utilizing Extensive Data Preprocessing Techniques

This research delves into the classification of cardiovascular disease (CVD) utilizing state-of-the-art machine learning algorithms, namely Support Vector Machine (SVM) and Artificial Neural Network (ANN). Before model training, extensive data preprocessing techniques were implemented, including data cleaning, feature scaling, encoding, Feature selection, handling imbalanced data, normalization, and cross-validation. After data preparation, an extensive evaluation of performance was carried out against various parameters like accuracy, precision, specificity, positive likelihood ratio (LR+), negative likelihood ratio (LR-), and diagnostic odd ratio (DOR). The comparison of SVM and ANN techniques indicates that the SVM has a better sensitivity in detecting positive cases while ANNs have more accuracy in the classification. This paper not only documents the use of new methods but also highlights the advantages and disadvantages of SVM and ANN models, and therefore helps to improve the use of machine learning applications in making health care decisions on CVD diagnosis.