Artificial Flora Algorithm-Based Feature Selection With Support Vector Machine for Cardiovascular Disease Classification

Accurately categorizing medical information is crucial for determining effective cardiac treatment options, especially as the volume of data grows and feature selection becomes increasingly challenging. This work proposes a model to identify the presence of Cardiovascular Disease based on various patient features, aiming to enhance prediction accuracy through a powerful feature selection method. This approach utilizes the Cleveland dataset by combining the Artificial Flora Optimization algorithm with the Support Vector Machine. The proposed algorithm functions as a meticulous gardener, selectively identifying the most significant features for heart disease prediction through an objective function. The model demonstrates impressive performance, achieving an accuracy of 96.63%, specificity of 95.73%, sensitivity of 97.74%, precision of 94.89%, and an F1-score of 96.29%. The model promises high-accuracy heart disease predictions by optimizing feature selection, potentially transforming clinical practice, and advancing research. The novel combination of the proposed technique holds significant potential for improving medical categorization and patient outcomes.