Early Prediction of Cardiac Arrest Using Hybrid Machine Learning Models

This study aims to predict cardiac arrest (CA) in real-world scenarios using machine learning (ML) classifiers. Previous research has shown that ML classifiers can effectively predict CA by analyzing various features. In this study, we applied multiple ML algorithms to a heart disease health indicator dataset that contains 21 features. We also proposed a hybrid model that combines Support Vector Machine (SVM) and Decision Tree (DT) algorithms. The results indicate that our hybrid model outperformed other ML classifiers with an accuracy of 91.56%. These findings suggest that our proposed model can be utilized for predicting CA. Additionally, this study highlights the potential of feature extraction techniques to improve the accuracy of CA prediction on a larger scale. Overall, our hybrid SVM-DT model appears to be superior to other independently used ML models, demonstrating the effectiveness of hybrid models in this context.