Context of data sharing practices in collaborative human genomic research in low and middle income countries: A systematic review.

Breastfeeding is globally recognized as the “gold standard” for infant nutrition, playing an irreplaceable role in building the infant immune system, promoting neurocognitive development, and reducing long-term disease risks for both mothers and infants [,]. Although the World Health Organization (WHO) and various national health departments strongly recommend exclusive breastfeeding for the first six months of life, global adherence remains suboptimal [,]. In China, the rate of exclusive breastfeeding often drops precipitously within three months postpartum due to multidimensional barriers such as return-to-work pressure, insufficient social support, and physiological health issues [–]. Therefore, exploring the key risk factors leading to early weaning is of urgent significance. Breastfeeding is a complex behavior regulated by multidimensional factors. Previous epidemiological studies indicate that sociodemographic characteristics [,], obstetric factors [,], and perinatal physiological indicators are fundamental variables. Among them, milk stasis and mastitis are widely considered the main physiological barriers [,]. However, traditional statistical studies often treat these factors as independent variables, employing linear models for analysis []. Recent studies point out that there are complex synergistic effects among risk factors which are difficult for traditional linear models to capture [,]. With the rise of medical big data, Machine Learning (ML) has become an important tool for clinical predictive modeling []. Multiple studies confirm that algorithms such as Random Forest and Support Vector Machines (SVM) significantly outperform traditional regression models in predicting breastfeeding outcomes [,]. Specifically, XGBoost provides a highly scalable and robust tree-boosting system capable of handling complex tabular data efficiently [].