Maternal Health Risk Detection using Stacking Classifier

Maternal health is a major cause for concern and the timely identification of health risks during pregnancy has been shown to greatly improve both the health condition of the mother and the baby. In this article, the prediction of maternal health risk using Stacking Classifier is proposed. The performance is evaluated using well-known classification performance evaluation metrics such as accuracy, recall, precision, F1-score, and Geometric Mean (GM). The performance of the Stacking Classifier is compared with other ten other Machine Learning (ML) models such as Logistic Regression (LR), K-Nearest Neighbors (KNN), Naïve Bayes (NB) Classifier, Support Vector Machine (SVM), Extreme Learning Machine (ELM), Decision Tree (DT), Random Forest (RF), Gradient Boosting (GB), eXtreme Gradient Boosting (XGBoost), and Adaptive Boosting (AdaBoost). The hyperparameters of all ML models are tuned using Random Search technique and all models are cross-validated using k-fold cross-validation technique. The Stacking Classifier has been found to outperforms other models.