Ứng dụng trí tuệ nhân tạo dự đoán tình trạng dinh dưỡng của thanh niên Việt Nam

Abstract

This study aims to develop and evaluate an artificial intelligence (AI) model for predicting the nutritional status of Vietnamese youth based on large-scale data. The dataset was derived from a nationwide survey involving 12,042 university students. Advanced machine learning algorithms, including XGBoost, LightGBM, and CatBoost, were applied to classify nutritional status using the Body Mass Index (BMI). Among these, the XGBoost model achieved the highest performance, with an accuracy of 75.3% and an F1-score of 73.75%. Key predictive variables included waist circumference, hip circumference, sleep duration, gender, late-night eating habits, fast food consumption, and daily dietary intake (energy, protein, fat, and micronutrients). The findings highlight the potential of AI in public health data analysis, particularly in identifying nutritional risks and generating personalized recommendations for diet and physical activity among youth. This research contributes to the development of precision nutrition approaches and supports the broader integration of digital health technologies in Vietnam’s public health strategies.