An Evaluation of Machine Learning Algorithms for an Enhanced Precision Healthcare in Stroke Prediction

Stroke poses a significant health challenge worldwide and ranks among the top causes of death and prolonged disability. Early prediction is essential to mitigate adverse outcomes, but existing predictive models often face challenges such as class imbalance and limited evaluation. This study aims to provide actionable insights into improving stroke prediction by systematically evaluating and comparing the performance of three widely used machine learning algorithms—Random Forest, Logistic Regression, and Support Vector Machines (SVM). The study addresses gaps in previous research by considering multiple evaluation metrics and integrating class balancing techniques such as the Synthetic Minority Over-sampling Technique (SMOTE). The dataset, comprising 10,091 records with demographic, clinical, and lifestyle attributes, was balanced using SMOTE. Random Forest achieved the highest performance with an accuracy of 98% and an AUC of 0.98, demonstrating its robustness and suitability for clinical integration. SVM also exhibited competitive performance, achieving an accuracy of 96% and an AUC of 0.96, while Logistic Regression showed limitations in recall (88%) and AUC (0.91). The findings underscore Random Forest’s potential as a reliable tool for stroke prediction and emphasise the importance of dataset balancing and comprehensive model evaluation. This study contributes to the advancement of predictive healthcare tools by providing a framework for selecting the most effective model for real-world stroke prediction applications. Keywords: Class Balancing, Clinical Decision-Support Systems, Feature Selection, SVN, Accuracy, Machine Learning, Predictive Modelling, Random Forest, Synthetic Data Proceedings Citation Format Fatimah Adamu-Fika, Deborah Ifeoluwa Ayeku, Tsentob Joy Samson, Aisha Tijjani Ramalan, Aanuoluwapo Enyojo Baba-Onoja, Oluwaseyi Ezekiel Olorunshola, & Henry Onyeoma Mafua (2024): An Evaluation of Machine Learning Algorithms for an Enhanced Precision Healthcare in Stroke Prediction. Proceedings of the 38th iSTEAMS Multidisciplinary Bespoke Conference. 17th – 19th July, 2024. University of Ghana, Accra, Ghana. Pp 318-329. dx.doi.org/10.22624/AIMS/ACCRABESPOKE2024P32