Prediction and Efficiency of Stroke Disease using data mining technique

Abstract

This research applies data mining techniques to compare the appropriate models. The predictions and efficiency of Stroke Disease. It was found that the significant factors influencing stroke disease included 10 factors consisting of performance focus on Gender, age, hypertension, heart disease, ever married, work type, residence type, avg glucose level, BMI, and Smoking Status. The model was used to compare 3 techniques: Decision Tree, Naïve Bayes, and K-Nearest Neighbors. The results showed that the K-Nearest Neighbors technique was the most suitable for predicting Stroke disease. By measuring the performance of the model with an Accuracy of 97.76%. Decision Tree performance with an accuracy of 97.09%. and Naïve Bays performance with an accuracy of 93.60%.