Stroke Detection Using Logistic Regression

Abstract

In this study, we present a stroke detection algorithm developed in Python, addressing a critical aspect of medical informatics. Drawing on a comprehensive public dataset, the detection program utilizes a logistic regression method to achieve an impressive accuracy rate of 94%. The dataset encompasses diverse demographic and health-related variables, such as marital status, age, BMI, heart disease history, working status, smoking habits, and glucose levels, which are all compared to whether or not that patient had a stroke.