An Extensive Machine Learning Framework Employing Ensemble Classifiers for Heart Disease Detection

Heart disease is a significant global health challenge, impacting many lives and often being a leading cause of mortality and illness. Detecting heart disease accurately and promptly is essential for effective medical care. Due to the complex interplay of risk factors and the intricate nature of diagnosis, advanced technologies like machine learning are being explored. This study introduces a comprehensive framework that employs machine learning techniques for heart disease detection. The framework involves critical stages such as data preprocessing, feature selection, and training ensemble classifiers. Experimental results highlight the importance of specific attributes in each classifier, offering insights into predictions. Notably, attributes like oldpeak, thalach, ca, cp, and thal are key features in the Random Forest classifier, while Gradient Boosting underscores ca, cp, thal, oldpeak, and age. AdaBoost and the Extra Trees Classifier reveal similar significant attributes. Evaluation metrics emphasize the models’ robust performance, with the Extra Trees Classifier achieving the highest accuracy. This research contributes to a practical heart disease detection framework with potential for improving medical decision-making.