STUNTING RISK PREDICTION METHOD IN PREGNANT WOMEN BASED ON ANTENATAL CARE DATA BASED ON ARTIFICIAL INTELLIGENCE

Abstract

This invention provides an ARTIFICIAL INTELLIGENCE BASED method for predicting the risk of stunting in pregnant women based on antenatal care (ANC) data. The method includes automatically grouping ANC data into datasets based on trimesters of pregnancy, forming machine learning prediction models as an application of artificial intelligence, independently for each trimester so that different prediction models are produced in each trimester, storing prediction results incrementally in a database without reprocessing or modifying the model or prediction results of the previous trimester when the next trimester data is received, and presenting prediction results in an interface that displays risk probability values, risk categories, and developments in risk changes between trimesters to support monitoring of stunting risks during pregnancy.In one implementation, the development of a predictive model is supported by data preprocessing, risk labeling, class distribution balancing, and machine learning model training. The method according to this invention enables continuous monitoring of stunting risk development and supports more timely intervention decision-making.