Comparative Study on Brainstroke Prediction Using Data Balancing Techniques

Brain stroke is a medical malady that occurs when there is an obstruction in the flow of blood in the brain or when a blood vessel in the brain ruptures, causing leakage. The global incidence of the ailment is steadily on the rise. Recent study indicates that identifying strokes at an early stage is the pivotal element in enhancing both patient care and treatment results. Timely prediction of brain stroke is crucial for implementing preventive measures and improving patient outcomes. Although machine learning algorithms are employed for automated diagnosis. However, these algorithms have specific constraints regarding accuracy. Another issue is that models trained on class-unbalanced datasets often have subpar results. Therefore, the primary aim of the proposed work is to include a pre-processing method for balancing the data before the machine learning algorithms to improve classification accuracy. we present a comprehensive comparative study on the application of data balancing techniques like SMOTE, ADASYN and hybrid technique like SMOTE+TOMEK links to enhance the performance of brain stroke prediction models. Our proposed methodologies outperform other arts-of-methods in terms of accuracy specificity, sensitivity, and other characteristics.