Penerapan Model Artificial Neural Networks (ANN) dalam Mengklasifikasi Risiko Kesehatan Ibu Hamil

Afrian, M.Alawi
Priyanto, Dadang
Sulistianingsih, Neny

Jurnal Sains dan Teknologi
Vol. 5 No. 3 (2025): September-Desember

This study employs an Artificial Neural Network (ANN) to classify maternal health risk levels using 29 medical variables. The objective is to develop a predictive model capable of identifying low-risk and high-risk pregnancy conditions as an early detection tool to support maternal health services. The dataset, obtained from Puskesmas Selong, underwent preprocessing steps including normalization, One-Hot Encoding, and class balancing using the SMOTE technique. The ANN architecture consists of three hidden layers equipped with ReLU activation, Batch Normalization, and Dropout, while model optimization is performed using the Adam optimizer and Focal Loss to address class imbalance. The model was trained using a 70%-30% train test split and evaluated through accuracy, precision, recall, and F1-score. The experimental results indicate strong model performance, achieving 97% accuracy, 98% precision, 99% recall, and 98% F1-score for the low risk class, as well as 90% precision, 81% recall, and 85% F1-score for the high risk class. The trained model was subsequently integrated into a web-based application, allowing users to input maternal health data and obtain automated risk predictions. These findings demonstrate that ANN can serve as an effective approach for supporting early maternal risk identification within AI-based clinical decision support systems.