Prediction of Brain Stroke using Machine Learning with Relief Algorithm

Stroke is one of the leading factors of fatality in people today. Numerous conditions, including stress, high blood pressure, cholesterol, obesity, type 2 diabetes, and dyslipidemia illnesses, may all contribute to stroke. By measuring the recorded values of the patients for about 31 features, such as heart rate, cholesterol level, blood pressure, heart rate, diabetes, metabolic syndrome, haemoglobin, steatosis, fibrosis, hepatocellular carcinoma, Body Mass Index (BMI) and so on. The proposed system uses machine learning to predict strokes in patients. The patient data is used to anticipate their health conditions and so as to save their lives. To anticipate strokes, comparing machine learning techniques K-Nearest Neighbour(KNN), Decision Tree (DT), Random Forest (RF), Support Vector Machine (SVM), and Logistic Regression (LR) are used. The relief algorithm is used to select the necessary attributes. An ensemble model comprising SVM, RF, and KNN gives better accuracy compared to base models.