| dc.description.abstract | Cardiovascular diseases (CVDs) are still a major global health concern, which
emphasizes the vital need for trustworthy classification models to facilitate early
detection and treatment. This is particularly important in areas where a shortage
of physicians, particularly cardiologists, makes it more difficult to diagnose and
treat patients with cardiac problems in a timely manner. This study uses strong
machine learning techniques to provide a comprehensive framework for heart
disease classification. This discovery represents a major step forward in
improving outcomes for CVDs, lowering mortality rates, and increasing access
to cardiovascular treatment through the use of web-based machine learning. In
order to accomplish this goal, the study makes use of an extensive dataset that
was gathered in 2018 from the Kurdistan Regional Government of Iraq's
Directorate of Health-Erbil. The collection is centered on the comprehensive
surgical Specialty Cardiac Center. A number of well-known machine learning
models, including decision trees, random forests, XGBoost, logistic regression,
and ensemble learning, are evaluated for their classifyive power using measures
like F1-Score, accuracy, precision, and recall. In addition, a sensitivity analysis
is carried out to investigate the effects of various train-test split ratios on model
efficacy, with the 80-20 ratio being determined to be the best option. Remarkably,
every model performs exceptionally well, with the exception of the decision tree,
random forest, XGBoost, and ensemble models. The user interface of the
Streamlit program is interactive and intuitive, making it easy for users to access
real-time classifications | en_US |
