Predictive Modeling of Brain Stroke Risk Using Supervised Machine Learning Techniques

By improving prediction technology and revealing links between many elements, artificial intelligence (AI) and machine learning (ML) have been profoundly changing the domains of biological and clinical research. This work investigates how several labelled factors could categorize brain stroke risk, therefore estimating the probability of a stroke for a patient. A brain stroke is a disturbance of blood flow to the brain that usually results in hemorrhagic disorders. With consideration for hypertension, obesity, cardiovascular disease, average blood sugar levels, smoking habits, past strokes, and age, this study intends to offer an early diagnosis of stroke risk by employing many machine-learning approaches. To investigate these elements, the work uses many supervised ML techniques like SVM, CatBoost, K-Nearest Neighbors (KNN), and Logistic Regression (LR). Particularly the CatBoost algorithm obtained the best accuracy at 93.67%. Early treatments and the adoption of a healthy lifestyle enabled by these precise forecasts help one to prevent strokes and advance long-term well-being.