Deep Learning Approach for Stunting Classification in Toddlers

Stunting in toddlers is a pressing global health issue that demands careful consideration. Stunting refers to a child's height falling significantly below the expected standard for their age, which can hinder physical and cognitive development and increase the risk of chronic diseases later in life. Early detection of stunting is crucial for timely intervention and minimizing long-term consequences. While previous studies have used machine learning to classify stunting in toddlers, their accuracy has been lacking. This research aims to improve classification accuracy using deep learning techniques. The study involves data collection, preprocessing data, and building a stunting classification model with Deep Neural Network. Results show a 3.42% improvement in training accuracy and 2.32% in testing accuracy compared to prior research. The developed model addresses overfitting/underfitting concerns, paving the way for tailored stunting classification tools for toddlers.