Prediction of the Occurrence of Stroke Based on Machine Learning Models

Abstract

The research conducted in the medical domain addressed a topic of significant importance, steadily growing in relevance each year. The study focused on predicting the onset of strokes, a condition posing a grave risk to individuals' health and lives. Utilizing a highly imbalanced dataset posed a challenge in developing machine learning models capable of effectively predicting stroke occurrences. Among the models examined, the Random Forest model demonstrated the most promising performance, achieving precision, recall, and F1-score metrics of 90%. These findings hold potential utility for healthcare professionals involved in stroke diagnosis and treatment.