Cardiovascular disease (CVD) remains a leading cause of mortality globally, necessitating early and accurate detection methods. This project presents a novel approach to CVD classification using an Artificial Flora Algorithm (AFA) for optimized feature selection combined with Support Vector Machine (SVM) for predictive classification. The system classifies the presence and severity of heart disease into four categories: normal, mild, moderate, and severe, based on clinical and behavioral inputs such as age, gender, chest pain type, fasting blood sugar, heart rate, and ST segment slope. To benchmark performance, multiple machine learning algorithms—Random Forest, Decision Tree, Multi-Layer Perceptron (MLP), XGBoost, and CatBoost classifiers—are employed and compared. The integration of AFA enables more accurate and relevant feature selection, thereby enhancing the model's generalization and reducing overfitting. Experimental results demonstrate that the Machine Learning model outperforms baseline models in terms of classification accuracy and efficiency. This research contributes to the development of intelligent, interpretable, and automated decision-support systems in cardiovascular healthcare.

Keywords: Cardiovascular Disease, Feature Selection, Artificial Flora Algorithm, Support Vector Machine, Machine Learning, Heart Disease Classification, XGBoost, CatBoost, MLP, Random Forest, Clinical Decision Support.