Abstract

Technology in the field of healthcare has been in constant enhancement, incorporating machine learning and various other technologies for clinical decision support systems or computer-aided healthcare systems that assist in examining complications. Cardiovascular disease is an issue of concern for any age group and gender and depends on innumerable factors like cholesterol, blood pressure, chest pain, and more. Furthermore, initial methods of getting a check-up for the doctor to determine whether the patient is suffering or will be suffering from the disease are time-consuming as well as laborious. Considering this, by using machine learning, it is possible to overcome the complications in order to produce an effective and accurate means of detecting cardiovascular diseases. In this work, the effectiveness of the popular ensemble learning methods (Bagging, Gradient Boosting, AdaBoost, Random Forests, and Extra Trees) is incorporated with Ant Colony Optimization (ACO) to predict cardiovascular diseases based on the UCI Cleveland dataset. The ensemble learning method Bootstrap Aggregation produced the highest accuracy of 95.08 when optimized with Ant Colony Optimization.