Enhancing the Accuracy of Heart Disease Prediction: A Comparative Study of Random Forest and Gradient Boosting Models Based on Physiological Indicators

Heart disease is a prevalent and critical health issue, making its prediction a major focus in medical research. However, traditional predictive models often face challenges such as low accuracy, poor model fitting, and limited robustness. To address these issues, this study improves upon traditional decision tree-based models by introducing two ensemble learning approaches: Random Forest and Gradient Boosting Tree. These methods significantly enhance prediction accuracy and effectively reduce the risk of overfitting. Using a dataset of over <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{1, 3 0 0}$</tex> patient records from the Kaggle platform, a series of data preprocessing techniques were applied to improve data stability. Experimental results demonstrate that the proposed methods achieve high accuracy and precision in heart disease risk prediction. This research provides a robust analytical tool for clinical decision support and early screening of heart disease, offering valuable potential for practical application.