Biomarker-Based Prediction for Monitoring the Risk of Preeclampsia Using Machine Learning Models

Preeclampsia is a severe hypertensive disorder of pregnancy. It endangers maternal and fetal health. Early identification is essential to reduce complications. Biomarkers offer objective signals of disease development. Accurate prediction supports timely clinical intervention. This paper provides a concise, academically oriented review of biomarkers used for the detection of preeclampsia. It then evaluates several machine learning (ML) algorithms, including Logistic Regression (LogReg), Random Forest (RF), Gradient Boosting (GB), Support Vector Machine (SVM), and Multi-Layer Perceptron (MLP), across multiple experimental scenarios designed to predict preeclampsia using the 'Proteomic Models in Preeclampsia' dataset. The findings indicate that Logistic Regression and SVM outperform the other ML models across standard classification metrics, underscoring their potential for predictive modeling in this clinical context.