Trends and Innovations in Stunting Prediction Using Machine Learning: A Bibliometric Analysis of Scopus Literature (2019-2024)

Abstract

Background: Stunting is a global malnutrition issue affecting children's development worldwide. Machine learning (ML) has been applied to predict stunting; however, studies on international collaboration and ML application across various geographical contexts remain limited. Objective: This study aims to analyze publication trends and collaboration patterns in stunting prediction research using machine learning and identify the most relevant ML methods. Methods: A bibliometric analysis was conducted using the Scopus database, focusing on publications from 2019 to 2024. The data were analyzed through descriptive statistics and science mapping, including trend topics, co-occurrence networks, and international collaboration analysis. Results: The study found that regression-based algorithms and deep learning are the most widely used machine learning methods for stunting prediction. International collaborations between countries with high stunting prevalence, such as Indonesia, Bangladesh, and Ethiopia, were also identified. Conclusion: This study highlights the importance of developing locally tailored predictive models and strengthening international collaboration to improve the effectiveness of stunting prediction models. Cross-sector and interdisciplinary collaborations are also essential for more holistic solutions.