TY - JOUR
AU - Dina Elly Yanti
AU - Lizah Framesti
AU - Anita Desiani
PY - 2023/08/22
Y2 - 2024/02/25
TI - PERBANDINGAN ALGORITMA C4.5 DAN SVM DALAM KLASIFIKASI PENYAKIT ANEMIA
JF - Jurnal Informatika Polinema
JA - JIP
VL - 9
IS - 4
SE - Articles
DO - 10.33795/jip.v9i4.1381
UR - http://jip.polinema.ac.id/ojs3/index.php/jip/article/view/1381
AB - Anemia is a condition when the hemoglobin in the body cannot function properly. Anemia has a bad impact on health, one of which is on the immune system. To prevent the occurrence of anemia can be done early detection by utilizing a mathematical approach using data mining. Data mining has classification methods that can be used for early detection of anemia. Methods that can be used for classification include the Support Vector Machines (SVM) algorithm and the C4.5 algorithm. This study applies the SVM algorithm and the C4.5 algorithm for the classification of early detection of anemia. The purpose of this study was to obtain the most appropriate method between the SVM algorithm and the C4.5 algorithm in the classification of anemia. This study applies percentage split testing techniques and k-fold cross validation. In the percentage split, a split of 80% was chosen as the training data and 20% as the test data. In k-fold cross validation, the value of k is chosen to be 10. The results of applying the two methods show that k-fold cross validation works better than percentage split with the percentage values for accuracy, precision, and recall being higher for each algorithm. For the performance of the two algorithms, C4.5 in its application works better with accuracy, precision, and recall values respectively, namely 99.29%, 98.7%, and 99.69% compared to the SVM algorithm. From the results obtained, it can be concluded that the C4.5 algorithm with the k-fold cross validation testing technique produces the best performance value for anemia classification compared to other testing algorithms and techniques.
ER -