Machine Learning-Based Algorithms for Determining C-Section among Mothers in Bangladesh

C-section prevalence has increased drastically over the past few decades across the globe. This growth has been caused by an array of factors, including maternal, socio-demographic, and institutional factors, and it is a global concern in both developed and developing countries. Therefore, the objective of this study is to identify relevant risk factors for the delivery type, and find a more accurate ML-based model for identifying cesarean women. The number of C-sections performed in the nation has increased to at least 45 percent in the two years prior to 2022. Because of this, we have used multiple logistic regression and machine learning algorithms to determine cesarean delivery and identify the socio-demographic risk factor among mothers in Bangladesh. Bivariate analysis results revealed that higher educated mothers and fathers, the richest family, overweight mothers, and hospital delivery had a higher percentage of cesarean babies. With an accuracy of 83.74%, NB (naive Bayes) outperforms the other five classifiers. We can get more precise information than accuracy from the ROC curve and the AUC. Depending on the AUC value, we can see that among all classifiers, Logistic Regression (LR) and Random Forest (RF) provide the most accurate classification for determining c-section. Our findings contribute to a better understanding of how to categorize C-section intentions among Bangladeshi women. The technique will be useful in identifying the women who are most likely to undergo a C-section in the healthcare system. As a result, the government can launch an effective public awareness campaign.