Predicting the occurrence of heart attacks using data mining

The appearance of mild symptoms may precede the occurrence of heart attacks, and sometimes they are sudden without any apparent symptoms. Therefore, it is necessary to intensify efforts to predict the occurrence of heart attacks. This paper aims to predict the chances of a heart attack by analysing a Kaggle dataset that includes 303 different instances. This data was collected from the records of patients from 14 clinics in the USA in 2020. Using Weka a popular open-source data mining and machine learning tool different statistical and mathematical methods were used to analyse data and attributes. The study applied different algorithms such as Hoeffding Tree, J48, Meta classifier, and PART Classifier to predict individuals' likelihood of heart attacks based on their traits. Hoeffding Tree outperformed compared to other classifiers. The results of this study can be applied in the real field to anticipate and prevent heart attacks by providing exceptional care and ensuring rapid therapeutic intervention for people with a higher chance of a heart attack.