Stroke Risk Prediction Model Using Machine Learning

Stroke is a disease that will occur when the blood flow to the brain is disrupted or diminished, causing the cell to die and it will lead to disability or even death. According to the WHO, stroke is the second deadliest disease in the world. Currently, the development of the Industrial Revolution 4.0 is collaborating with the health sector by using Machine Learning to make an accurate prediction of stroke that is highly valuable for early intervention and treatment. This research uses machine learning to predict a patient who is at high risk to get a stroke. The research will use three machine learning algorithm models, including Naive Bayes, Decision Tree, and Random Forest. The prediction use patient health history as the attribute in each mode. After testing for each of these methods. The first method that has the best accuracy is the Random Forest method with 94,781% accuracy, followed by the Decision Tree method with 91,906% accuracy and the least accurate is the Naïve Bayes Method with 89,976% accuracy. Based on these results, it can be concluded that the Random Forest method has the highest accuracy than other methods.