Machine Learning-Based Analysis of Factors Influencing the Incidence of Heart Disease in Health Conditions

Heart disease is a cardiovascular disease that affects the everyday life of human beings and has a high mortality rate, and nowadays people do not pay attention to their health, which makes more and more people suffer from heart disease. Therefore, this study aims to analyze the importance of health factors that affect the risk of heart disease. This study analyzed the heart disease dataset using three machine learning methods Random Forest, Decision Tree, and eXtreme Gradient Boosting (XGBoost). The analysis yielded, the accuracy and confusion matrix of the three machine learning models, with Random Forest and XGBoost performing well in model fitting and Decision Tree performing slightly worse. The feature importance distributions analyzed revealed that age category, stroke, and difficultly walking were the three most important health factors affecting the risk of developing heart disease. Among them, the age category performs best in the decision tree, difficult walking has a very high importance in Random Forest and XGBoost, and stroke has a higher importance in all three models. This conclusion will be beneficial for medical diagnosis, doctors to propose more targeted diagnoses and treatment plans for patients' health conditions. It will benefit public health organizations to make more informed proposals to improve citizens' health awareness.