Healthcare with emphasis on coronary thrombosis prediction.

Heart attacks, also known as cardiac arrests, have become a leading cause of death globally in recent decades. They represent the culmination of intricate relationships between modifiable and non-modifiable risk factors. Many cases of cardiovascular disease can be attributed to factors that can be changed, making prevention possible for most cases. This research aims to develop a predictive model for the likelihood of patients experiencing a heart attack using pre-existing datasets from the UCI Heart repository database. In the study, classifiers were employed in a pipeline approach to machine learning, performing predictions in both directions, with and without optimizations and feature transformations. The findings indicate that the Random Forest classifier achieved the highest accuracy score in binary prediction, where 1 signifies a possibility of a heart attack and 0 denotes no chance. Factors such as age, cholesterol level (with levels above 200 mg/dl being more susceptible), increased heart rate, and the type of chest pain (with typical angina being the most common and asymptomatic chest pain being the least) were found to have the most significant influence on the prediction model. [ABSTRACT FROM AUTHOR]