Machine Learning Techniques for Maternal Health Risk Prediction: A Comparative Performance Analysis with Real-World Data

This study aims to identify the most effective predictive computational model in assessing maternal health risks. This was achieved through a comprehensive investigation involving the collection, application, and evaluation of various advanced data analysis methodologies. These approaches were integral to the development of the system. The data used in this crucial study was obtained from the UCI Machine Learning Repository, a rich resource of health information. The data included crucial features relevant to maternal health conditions. These features were selected because of their high relevance in indicating health risks. The results of a thorough evaluation showed that the model optimized through cross-validation (Ford) technique consistently performed superior to other approaches. The model achieved an impressive accuracy of 86.7%, precision of 87.0%, recall of 87.6%, and F1 score of 87.2%. Ultimately, this study highlights the significant potential of utilizing artificial intelligence to improve health risk prediction in pregnant individuals. Implementation of these findings may facilitate early detection and support more informed clinical decision-making, contributing to improved maternal health outcomes.