<![CDATA[Heart disease is one of the deadliest diseases worldwide. This condition often presents symptoms that do not immediately cause severe effects on the sufferer, making early anticipation crucial. To reduce fatalities caused by heart disease or cardiovascular disorders, a system is required to identify its primary causes so that these factors can be minimized. Therefore, this study applies the Chi-square feature selection method to determine the key features influencing the accuracy of Machine Learning models. A comparison is conducted between K-Nearest Neighbor, Naïve Bayes, Logistic Regression, Support Vector Machine, and Random Forest algorithms. This comparison aims to obtain the most accurate results, as a higher algorithm accuracy leads to a more precise classification system for heart disease. The study’s findings indicate that eight key features selected using the Chi-square method yield the highest accuracy, specifically 93.51% with the KNN algorithm. These results demonstrate that using relevant features improves classification accuracy and system efficiency compared to utilizing all available features. Consequently, this research contributes to the selection of essential features in Machine Learning algorithms through the Chi-square technique, ensuring a more effective and optimized heart disease classification system.]]>
Copyrights © 2025
