IDENTIFYING PREGNANCY LOSS RISK FACTORS USING MACHINE LEARNING

Abstract

Pregnancy loss, or spontaneous abortion, affects 15–20% of clinically diagnosed pregnancies. This study uses cross-sectional data from the Bureau of Statistics Punjab to identify key risk factors through various machine learning algorithms, including Logistic Regression, KNN, LDA, SVM, and others. Among these, KNN achieved the highest accuracy at 91%, while most models exceeded 80%. Feature selection methods revealed that "total children ever born" and "place of delivery" were the most influential factors. These findings highlight the potential of ML models in predicting pregnancy loss and identifying critical risk indicators.