Sistem Klasifikasi Risiko Kehamilan Dengan Algoritma CART (Classification And Regression Tree)
Zafira Farhani, Ir. Adhistya Erna Permanasari, S.T., M.T., Ph.D. ; Dr.Eng. Silmi Fauziati, S.T., M.T.
2024 | Skripsi | S1 TEKNIK BIOMEDIS
All pregnancies inherently have risks. However, there are certain conditions that can lead to a high-risk pregnancy. A high-risk pregnancy is a situation in which a pregnant woman has a pregnancy with abnormal health conditions that can cause health complications and endanger the safety of both the mother and the fetus. Many factors increase the risk of pregnancy. One of them is the delayed diagnosis of the pregnant woman's health condition. Early pregnancy diagnosis can help pregnant women reduce the risk of complications. By classifying pregnancy risks, pregnant women can take preventive actions to avoid health complications. This system is developed based on machine learning as its classification model. The machine learning method used is the decision tree algorithm, namely CART (Classification and Regression Tree). Before performing the classification with CART, the dataset is preprocessed using SMOTE to balance the number of data in each class. To improve accuracy performance, the results of the decision tree of the CART algorithm are pruned with CCP (Cost-Complexity Pruning). The accuracy result obtained from this classification is 90.57%. Meanwhile, the result without SMOTE and pruning is 83.74%. From the results obtained, the accuracy performance of the CART algorithm can be improved by implementing both techniques.
Kata Kunci : Risiko Kehamilan, Machine Learning, Decision Tree, CART (Classification and Regression Tree), CDSS (Clinical Decision Support System)