Machine Learning Model for Stroke Disease Classification

In many nations, stroke is the primary factor contributing to mortality and obesity. This study improves the quality of the picture to strengthen image results and to diminish noise to increase the image quality of CT scans of stroke patients, as well as using machine learning algorithms to categorize the scans of the patients into the two subtypes those are ischemic stroke and stroke hemorrhage. This research utilized eight machine learning algorithms to classify stroke diseases, those are Naive Bayes, K-Nearest Neighbors, Decision Tree, Logistic Regression, Multi-layer Perceptron (MLP-NN), Random Forest, Support Vector Machine and Deep Learning. Our study showed that Random Forest delivers the most accurate outcomes (95.97%), together with recall values (96.12%), f1- Measures (95.39%), and precision values (94.39%).