Stroke Disease Analysis and Classification Using Decision Tree and Random Forest Methods

Abstract

A stroke is a medical emergency that occurs when blood flow to the brain is blocked or decreased, depriving brain tissue of oxygen and nutrients. Stroke is the world's second leading cause of death, according to the World Health Organization (WHO). Stroke patients die within the first year of their illness. To reduce the risk of stroke, simple and effective tools are required. The goal of this study was to look into the classification of stroke potential and come up with a simple and reliable model. The Kaggle database provided the stroke prediction data set, which was based on input criteria such as gender, age, various illnesses, and smoking status. To determine the prediction of the construction model, decision trees and random forest classification methods were utilized. The independent variables determining the incidence of stroke were determined to be age (AUC 0.85), hypertension (AUC 0.62), blood sugar level (AUC 0.61), history of heart disease (0.56), married status (0.60), and body mass index (BMI) (AUC 0.56). Age, hypertension, blood sugar level, and BMI were all valid, with a random forest method accuracy of 98.90 percent and decision tree method accuracy of 95.90 percent.