<![CDATA[Stroke is a deadly disease. This can occur due to disturbances in brain function that occur suddenly, progressively and quickly. However, it is difficult to know the early symptoms of stroke. The application of data mining knowledge can be used to diagnose disease. This research was conducted to implement data mining in classifying brain stroke. The dataset used was obtained from Kaggle, totaling 4891 data. However, the dataset does not have a balanced amount of data for each class. To balance the data, the SMOTE technique is used which aims to increase accuracy. The application of the classification algorithms used, namely the Logistic Regression (LR), Random Forest (RF), Support Vector Machine (SVM), and K-Nearest Neighbors (KNN) algorithms aims to determine the best algorithm performance. This research resulted in a comparison of the four algorithms which showed that the LR, RF and SVM algorithms produced the highest accuracy, precision, recall and f1-score values, namely 95% accuracy, 95% precision, 100% recall and 97% f1-score. The KNN algorithm produces lower accuracy, precision, recall and f1-score values, namely 90% accuracy, 95% precision, 85% recall and 90% f1-score.]]>
Copyrights © 2024
