<![CDATA[Status gizi balita merupakan indikator penting dalam menilai kesehatan masyarakat dan kualitas sumber daya manusia di masa depan. Klasifikasi status gizi yang masih dilakukan secara manual berpotensi menimbulkan kesalahan dan ketidakkonsistenan, sehingga diperlukan pendekatan berbasis data. Penelitian ini bertujuan menerapkan algoritma Random Forest untuk mengklasifikasikan status gizi balita berdasarkan indikator antropometri. Data yang digunakan berasal dari Puskesmas Merancang Ulu, Kabupaten Berau, sebanyak 321 data balita tahun 2024. Setelah melalui tahapan seleksi fitur dan prapemrosesan data, diperoleh 18 atribut yang digunakan dalam pemodelan. Proses klasifikasi dilakukan menggunakan bahasa pemrograman Python dengan pustaka scikit-learn dan divalidasi menggunakan metode K-Fold Cross Validation. Kinerja model dievaluasi menggunakan Confusion Matrix dengan metrik akurasi, presisi, recall, dan F1-Score, serta ROC-AUC rata-rata makro. Hasil penelitian menunjukkan bahwa algoritma Random Forest menghasilkan akurasi rata-rata sebesar 88%, dengan nilai presisi, recall, dan F1-Score masing-masing sebesar 88%, serta nilai ROC-AUC rata-rata makro sebesar 0,965. Temuan ini menunjukkan bahwa Random Forest efektif dan andal sebagai sistem pendukung keputusan untuk klasifikasi status gizi balita di tingkat layanan kesehatan dasar. Nutritional status of children under five is a critical indicator for assessing public health and future human resource quality. Manual classification of nutritional status is prone to errors and inconsistency, highlighting the need for data-driven approaches. This study aims to apply the Random Forest algorithm to classify the nutritional status of under-five children based on anthropometric indicators. The dataset consists of 321 nutritional records of children collected in 2024 from Merancang Ulu Public Health Center, Berau Regency. After feature selection and data preprocessing, 18 attributes were used for model development. The classification model was implemented using Python with the scikit-learn library and validated through K-Fold Cross Validation. Model performance was evaluated using a Confusion Matrix with accuracy, precision, recall, and F1-score metrics, along with macro-average ROC-AUC. The results indicate that the Random Forest model achieved an average accuracy of 88%, with precision, recall, and F1-score each reaching 88%. Furthermore, the macro-average ROC-AUC value of 0.965 demonstrates the model’s strong capability in distinguishing multiclass nutritional status categories. These findings suggest that the Random Forest algorithm is effective and reliable as a decision support system for classifying the nutritional status of children under five in primary healthcare settings.]]>
