Application of the Supervised Learning Algorithm for Classification of Pregnancy Risk Levels

Abstract

MMR is the number of women who die due to disorders during pregnancy or their treatment (excluding accidents, suicides, or incidental cases) during pregnancy, childbirth, and during the puerperium or 42 days after giving birth. This research aims to classify pregnancy risk datasets, namely to compare the performance of the NBC, K-NN, and SVM methods on the pregnancy risk status dataset and to find out the accuracy comparison of the algorithm results above. From the results of the analysis, it was found that of the three algorithms it resulted in a classification of pregnancy risk levels with the highest value occurring at a high level. To determine the accuracy of the data, a comparison was made between the three algorithms. Based on the confusion matrix namely Accuracy, Precision, and Recall. The results of the comparison can be concluded that the KNN algorithm provides the highest accuracy of 77.55%, NBC of 69.39%, and the lowest accuracy by SVM of 67.35%. These results state that the KNN algorithm classifies pregnancy risk level data better than the other two algorithms