-
1
A novel scoring system and prediction algorithm were developed to identify infants at risk of stunting using growth, nutritional, and biochemical indicators.
-
2
The study involved 380 infants aged 0–12 months, with a training set of 266 and a validation set of 114, ensuring robust model evaluation.
-
3
Five significant indicators of stunting were identified, with length growth velocity being the strongest predictor among them.
-
4
The Gradient Boosting model outperformed others, achieving an AUC of 0.861 in the training set and 0.850 in the validation set.
-
5
A visual nomogram was created to facilitate quantitative risk assessment for stunting, enhancing clinical decision-making.
