<![CDATA[The condition of a baby's birth is an important indicator in assessing the health risks of mothers and children. This study aims to develop a model for predicting the risks of childbirth using the Naïve Bayes algorithm with the Cross Industry Standard Process for Data Mining (CRISP-DM) approach, which includes the stages of Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation, and Deployment. The dataset used consists of 634 pregnant women data obtained from the Bungbulang Community Health Center, Garut Regency. This study tested three variations of data preprocessing, namely One Hot Encoding, Label Encoding, and Min-Max Scaling. The evaluation results show that all Naïve Bayes models perform well with an accuracy above 94%. Among the three, the model with Min-Max Scaling produces the most optimal performance with an accuracy of 95.4%, precision of 95.7%, recall of 94.5%, F1-Score of 95.0%, and AUC reaching 100%. These findings indicate that the application of Min-Max Scaling to the Naïve Bayes algorithm is effective in improving prediction performance while providing balance in evaluation metrics. The results of this study are expected to support early identification of birth risks and contribute to decision-making in maternal and child health services.]]>
Copyrights © 2026
