Comparison of Machine Learning Models for Predicting Maternal Health Risk

To perform a comparative analysis of machine learning models for estimating maternal risk levels during pregnancy.

Machine learning is used for risk assessment and pregnancy complications; this project focuses on comparative accuracy and interpretability.

Early identification of maternal health risks, enabling timely interventions and improved resource distribution.

—This study presents a comparative analysis of machine learning models for providing an estimate about the maternal risk levels during pregnancy is presented in this particular study. XGBoost, Random Forests, Decision Trees, Gradient Boosting, and Logistic Regression are the models which were used for assessment. These models were trained and tested on a dataset of 1014 entries which included characteristics such as age, blood pressure, blood sugar, body temperature, and heart rate. The accuracy of XGBoost (84.24%) is better than the other models. According to the results, it is closely followed by the Decision Trees Model (81.77%). Decision Trees offer better interpretability, which is essential for clinical decision making, even after XGBoost provides a better predictive performance. By offering individualized risk assessments, this study demonstrates how machine learning can be used to improve maternal health outcomes. It is also crucial to strike a balance between model accuracy, interpretability and practicality which this research also emphasizes on. In order to improve the predictive capabilities, future research directions will examine the hybrid models and incorporate extra clinical features.