Performance Enhancement of Machine Learning Algorithms for Predicting Stroke

Stroke is very common heath problem now. Early prediction of stroke is very essential to decrease the impact of this disease. Machine Learning (ML) algorithms are very useful in making predictions for various diseases. In this paper, the power of machine learning algorithms is used for stroke prediction. Even after a lot of research the accuracy of machine learning algorithms is a challenging issue. Many researchers worked on improving the accuracy of ML models. Machine learning models worked on data sets. For stroke prediction using machine learning, a well-known data set is considered in this paper. The data set consist of many important factors which may affect the chances of stroke in a person. Logistic Regression, Decision Tree and Random Forest are three machine learning algorithms which are very popular and effective in prediction of stroke. In this work these three machine learning algorithms are applied for stroke prediction and their performance is checked and compared. From the results it is observed that the performance of Logistic Regression is best with 95.34% accuracy.