An Efficient Fetal Health Prediction System Using Enhanced XGB Classifier Compared Over K-Nearest Neighbor Classifier with Less Complexity and Improved Accuracy

The aim of this research is to improve the Fetal Health Prediction using the XGB Classifier and in comparison, with K- Nearest Neighbor. Prenatal diagnostics, which includes screening and diagnostic testing, can give important details about the fetus's health and help people comprehend the risk. This study aims to offer clinicians and expectant mothers convenient services by gathering secondary data from diverse sources. In this study, a total of 3200 data points was utilized for our research purposes. Out of these, 2560 data points were allocated for training, while the remaining 640 data points were set aside for testing. The entire process was iterated 10 times for robustness. The study involved two distinct groups, with each group consisting of 10 samples. The K-Nearest Neighbor (KNN) Algorithm was categorized under Group 2, while the XGB Classifier Algorithm was assigned to Group 1. The statistical power of our analysis was set at 80%. The XGB Classifier Algorithm demonstrated a notably higher accuracy rate of 95.96% compared to the KNN Algorithm, which achieved an accuracy of 87.74%. The calculated significance value was 0.001 (p < 0.05), indicating that there is a statistically significant difference between the two algorithms considered for the development of the Novel Fetal Health Prediction System. In conclusion, our findings highlight that the XGB Classifier Algorithm outperforms the K-Nearest Neighbor Algorithm with an impressive accuracy rate of 95.96%, which is significantly higher than the 87.74% accuracy achieved by the latter. This underscores the potential superiority of the XGB Classifier Algorithm for the task at hand in the context of fetal health prediction.