AI-Driven Heart Disease Diagnosis: A Hybrid Model of Machine Learning Techniques

This study presents an intelligent heart disease diagnostic model that combines data normalization, Principal Component Analysis (PCA), DBSCAN clustering, and Random Forest classification to enhance diagnostic accuracy. Using a dataset from Zheen Hospital in Erbil, Iraq, with ۱,۳۱۹ samples and ۹ features, the model achieves an overall test accuracy of ۹۸.۱۱%, along with precision ۹۸.۱۷%, recall ۹۸.۷۷%, and F۱ score ۹۸.۴۷%. These results indicate that the model significantly outperforms traditional methods and state-of-the-art models in heart disease diagnosis. Nevertheless, the model may have limitations regarding generalizability to diverse populations and data quality. Future improvements could include enhancing the dataset and exploring additional machine-learning techniques. The integration of AI and machine learning in this paper adds a modern dimension to diagnostic tools used in the clinical setting.

کلیدواژه ها:

Heart Disease Diagnosis ، Machine Learning ، DBSCAN ، Random Forest ، Healthcare AI