Table 1 Summary of machine learning approaches for stroke prediction
Study | Models used | Accuracy score | Importance |
|---|---|---|---|
[15] | Naïve Bayes, Logistic Regression, XgBoost, SVM | SVM: 98.6% | SVM achieved the highest accuracy and precision (99.9%), highlighting its robustness. |
[16] | Various ML classifiers, Logistic Regression | AUC: 0.694 to 0.823 | Inclusion of SDoH features significantly improved AUC, showing the importance of these variables. |
[17] | Logistic Regression with RFE | AUC:>0.94 | Recursive feature selection and follow-up data incorporation enhanced predictive utility. |
[18] | Stacking (Naïve Bayes, Random Forests, LR) | AUC: 98.9%, Accuracy: 98% | Stacking method demonstrated superior performance across multiple metrics. |
[19] | Random Forest, XGBoost | Random Forest: 90.36%, XGBoost: 89.02% | SHAP and LIME techniques enhanced interpretability, with Random Forest performing best. |
[23] | XGBoost | Highest Accuracy | XGBoost showed superior predictive power for pre-screening ischemic stroke. |
[24] | DeepSurv, Deep-Survival-Machines | Enhanced Predictive Accuracy | Deep learning models surpassed traditional survival models for predicting MACEs after AIS. |