A Comparative of Classification Models for Predicting of Heart Diseases

To slow the rising mortality toll, early detection of heart illness is essential, as it is one crucial organ in human; if it stops beating, life ends. To put it simply, cardiovascular diseases (CVDs) are the important global killer. In 2019, CVDs were responsible for the deaths of 17.9 million individuals worldwide. The medical field has benefited greatly from technological advancements, but there are still some challenges associated with machine learning, such as data that is noisy, error-prone, or contains outliers. Due to their often-late diagnosis, these diseases are principal reason of bereavement in heart patients, making it imperative to learn more about effective algorithms for disease prognosis. Accurate predictions of such results with Machine Learning algorithms can support in the initial diagnosis of diseases, mitigation of their effects, and even prevention of their occurrence by providing crucial information to both medical professionals and patients. This study proposes a model that can rapidly and cheaply diagnose cardiac problems. Therefore, this research employs SVM, Naive Bayes, and Random Forest to detect cardiac illnesses using a data set from Kaggle, with 93%, 85%, and 95% accuracy, respectively; however, the Random Forest attained the best accuracy of 96.9%.