KLASIFIKASI STUNTING PADA BALITA MENGGUNAKAN ALGORITMA DECISION TREE PADA MACHINE LEARNING

La Denna Hasri Monasari; Nunik Pratiwi;

Klasifikasi Stunting Pada Balita Menggunakan Algoritma

Decision Tree Pada Machine Learning
La Denna Hasri Monasari

Stunting adalah gangguan pertumbuhan yang dialami balita akibat kekurangan gizi sehingga
menghambat pertumbuhan fisik dan perkembangan kognitif. Faktor penyebabnya meliputi
kurangnya asupan nutrisi, rendahnya pendidikan orang tua, serta lingkungan kesehatan yang tidak
memadai. Penelitian ini bertujuan untuk mengembangkan model klasifikasi status gizi balita guna
mendeteksi risiko stunting secara dini dengan menerapkan algoritma Decision Tree. Model
mengkategorikan status gizi balita berdasarkan fitur usia (bulan), jenis kelamin, dan tinggi badan
(cm). Data diperoleh dari platform Kaggle, terdiri atas 121.000 data. Tahapan penelitian mencakup
proses pengumpulan data, pre-processing data (cleaning data, deteksi outlier, dan label encoding),
eksplorasi data, pemisahan data latih dan uji, pelatihan model, serta evaluasi performa. Evaluasi
dilakukan menggunakan confusion matrix, yaitu matriks evaluasi yang menunjukkan jumlah
prediksi benar dan salah pada masing-masing kelas. Hasil evaluasi model Decision Tree
memperoleh akurasi sebesar 95.73% dalam mengklasifikasi status gizi ke dalam empat kategori
yaitu normal, stunted, severely stunted, dan tinggi. Temuan menunjukkan bahwa algoritma Decision
Tree efektif dalam mengidentifikasi kondisi stunting serta mendukung sistem pemantauan gizi anak
balita secara digital.
Kata kunci: balita, decision tree, klasifikasi, machine learning, stunting, status gizi

Classification of Stunting in Toddlers Using the Decision Tree

Algorithm in Machine Learning
La Denna Hasri Monasari

Stunting is a growth disorder experienced by toddlers due to malnutrition, which hinders physical
growth and cognitive development. The causes include insufficient nutritional intake, low parental
education, and inadequate health environments. This study aims to develop a nutritional status
classification model for infants to detect the risk of stunting early by applying a Decision Tree
algorithm. The model categorizes infants' nutritional status based on age (months), gender, and
height (cm). The data was obtained from the Kaggle platform, consisting of 121,000 data points.
The research stages include data collection, data pre-processing (data cleaning, outlier detection,
and label encoding), data exploration, training and testing data separation, model training, and
performance evaluation. Evaluation was conducted using a confusion matrix, which is an evaluation
matrix showing the number of correct and incorrect predictions in each class. The evaluation results
of the Decision Tree model achieved an accuracy of 95.73% in classifying nutritional status into
four categories: normal, stunted, severely stunted, and high. The findings indicate that the Decision
Tree algorithm is effective in identifying stunting conditions and supports a digital monitoring
system for the nutrition of infants and toddlers.
Keywords: toddler, decision tree, classification, machine learning, stunting, nutritional status