Stroke Risk Prediction Using Machine Learning Algorithms
Rishabh Gurjar, H K Sahana, C Neelambika, Sparsha B Sathish, S Ramys
Abstract
The majority of strokes are brought on by unforeseen obstruction of pathways by the heart and brain. Distinct classifiers have been developed for early detection of different stroke warning symptoms, including Logistics Regression, Decision Tree, KNN, Random Forest, and Naïve Bayes. Furthermore, the proposed research has obtained an accuracy of around 95.4%, with the Random Forest outperforming the other classifiers. This model has the highest stroke prediction accuracy. Therefore, Random Forest is almost the perfect classifier for foretelling stroke, which doctors and patients can utilise to prescribe and identify likely strokes early. Here in our research we have created a website to which model is dumped/loaded, such that the interface will be friendly to the end-users.