<![CDATA[The total number of deaths worldwide due to heart failure continues to show an increase. Classifying patients with the best accuracy can help improve preventive measures based on clinical information. This study compares classification algorithms including C4.5, K-Nearest Neighbor, and Naïve Bayes based on CRISP-DM with the 10-fold cross-validation model evaluation technique and pairwise t-test using RapidMiner software. The research obtained the highest accuracy value of 0.779 with a standard deviation of approximately 0.046.. The research results indicate that the C4.5 algorithm performs the best, followed by the Naïve Bayes algorithm with a statistically insignificant difference, and lastly, the K-Nearest Neighbor algorithm with the smallest value, thus considered less suitable for implementation in the dataset.]]>
Copyrights © 2024
