<![CDATA[Ischemic Heart Disease (IHD) is the leading cause of death worldwide, accounting for 13% of global fatalities. The number of deaths caused by IHD rose from 2.7 million in 2000 to 9.1 million in 2021, an increase of 6.4 million. IHD can be diagnosed through medical examinations or various health tests, as well as by leveraging technological advancements in artificial intelligence to enable early disease detection. This early detection is crucial for preventing heart disease, as there is currently no cure for the condition. This study aims to compare machine learning algorithms based on decision tree methods (Decision Tree, Random Forest, and Gradient Boosted Tree) with optimization using genetic algorithms to predict heart disease. The dataset used includes information from 8,625 patients who have experienced heart attacks, featuring attributes such as Sex, General Health, Age Category, Height (in meters), Weight (in kilograms), BMI, and "Had Heart Attack" as the label attribute. The initial modeling phase involved splitting the data into 80% for training (6,900 samples) and 20% for testing (1,725 samples). The results showed that the Random Forest model achieved the highest accuracy at 95.26%, narrowly surpassing the Decision Tree model, which attained 95.22%, by 0.04%. Meanwhile, the Gradient Boosted Tree model demonstrated the lowest accuracy at 90.99%. Subsequently, the application of the Genetic Algorithm significantly improved the accuracy, precision, and recall metrics across all three models, although the recall value for the Gradient Boosted Tree model decreased by 5.17%.]]>
Copyrights © 2025
