سامانه همایش های دانشگاه آزاد اسلامی واحد نجف آباد

صفحه اصلی / International Conference on Artificial Intelligence; City, Industry and Health

Long-term Prediction of Coronary Artery Disease Using Logistic Regression

نویسندگان :

Koosha Mardasi¹ Nasim Noorafza²

1- Department of Computer Engineering, Na.C, Islamic Azad University, Najafabad, Iran 2- Department of Computer Engineering, Na.C, Islamic Azad University, Najafabad, Iran

کلمات کلیدی :

cardiovascular،coronary artery disease،logistic regression،naive bayes،machine learning

چکیده :

This study addresses long-term coronary artery disease(CAD) prediction by presenting a noninvasive, machine learning framework that primarily employs Logistic Regression, with a comparative analysis against Naive Bayes. Utilizing extensive clinical datasets, our approach begins with rigorous data preprocessing applying median imputation for missing values and filtering individuals aged in a specific range. A dual feature selection strategy, combining mutual information analysis with an XGBoost-based relevance ranking, is implemented to isolate the most predictive variables. To counteract class imbalance, the SMOTEENN algorithm is applied alongside stratified sampling and normalization. The logistic regression model is optimized using grid search and cross-validation, ensuring robust performance. Comparative evaluations demonstrate that, while Naive Bayes offers a simpler model with certain advantages, Logistic Regression provides superior interpretability and clinical relevance, allowing for clearer assessment of individual risk factors such as smoking. Overall, the proposed framework offers a cost-effective and reliable tool for early CAD detection and risk stratification, thereby supporting enhanced clinical decision-making in precision cardiovascular medicine