Enhancing stroke disease classification through machine learning models via a novel voting system by feature...

Area of Science:

• Cardiology
• Machine Learning
• Data Science

Background:

• Heart disease is a major global health concern, driving the need for improved predictive diagnostics.
• Existing machine learning models for heart disease prediction often lack sufficient accuracy for clinical application.

Purpose of the Study:

• To develop and evaluate advanced machine learning models for accurate heart disease prediction.
• To enhance early detection and intervention strategies for cardiovascular diseases.

Main Methods:

• Applied nine machine learning algorithms: XGBoost, logistic regression, decision tree, random forest, k-nearest neighbors (KNN), support vector machine (SVM), Gaussian Naive Bayes (NB Gaussian), adaptive boosting, and linear regression.
• Utilized feature selection, grid search hyperparameter tuning, and cross-validation to optimize model performance and interpretability.
• Developed a novel voting system combined with feature selection for improved heart disease classification.

Main Results:

• XGBoost demonstrated superior performance, achieving 99% accuracy, precision, and F1-score, with 98% recall and 100% ROC AUC.
• Evaluated models using accuracy, precision, recall, F1-score, and ROC AUC to ensure comprehensive performance assessment.

Conclusions:

• The developed XGBoost model offers a highly accurate and reliable approach for early heart disease diagnosis.
• This study provides a significant advancement in predictive modeling for preventive cardiovascular healthcare.