Heart attacks are one of the most common diseases that increase mortality among cardiovascular patients, leading to a significant number of deaths worldwide. Each year's cases motivate the healthcare field to create and develop classification support software or tools to help with heart disease diagnosis. Machine learning is a powerful tool that can be used in the battle against cardiovascular diseases. The prognosis of heart diseases can be enhanced by analytical prediction to identify the risk level of patients and the required medical care at an early stage. This research developed several machine learning models, logistic regression, classification tree, naïve Bayes, and random forest to classify and predict heart diseases based on the patient's historical health-related data. Principal components analysis was also conducted to reduce the dimensionality of the independent variables and identify the most influential variables for the risk of heart disease. The study shows that the classification tree model provides the best results with 98.5% accuracy, 97.2% precision, and the highest area under the curve (AUC) level, 98.5%, compared to other machine learning models. As per the classification tree results, the three most essential features significantly impacting heart disease are the patient's age, emotional stress, and cholesterol levels. Additionally, the features have been ranked based on their effect on heart disease. Logistic regression sigmoid function was identified to estimate the probability of heart disease occurring. Such models provide healthcare and teams with a valuable classification tool to predict and categorize the patient's heart condition.
Sliti et al. (Tue,) studied this question.