Analysis of Stroke Classification Using Random Forest Method

Stroke is a disease in which the sufferer experiences or rupture of blood vessel brain so that does not get supply provides oxygen. Patients who suffer from stroke will experience cognitive disorders ranging decreased consciousness, visuospatial disorders, non-verbal learning communication and reduced levels patient attention. Data World Organization shows there are 13.7 million new cases every year, about 5.5 deaths occur due to stroke. This research aims analyze attributes any variables affect classification strike test performance form accuracy, precision, recall, f-measure. The method used random forest using tree, namely 50, 100, 200, 500. divided into no data 5110, 70% training 30% testing data. results showed 100 trees was better than 500 trees, with an accuracy value 86.82%, precision 15.76%, recall 38.15%, f1-score 22.30% after doing SMOTE.