Performance Comparison of Machine Learning Algorithms to Predict Labor Complications and Birth Defects Based On Stress

Stress affects physical as well as the mental health of the people and it follows that the stress is one of the major reasons behind the complications during pregnancy like hypertension. Hence it is necessary to ascertain the effects of stress on the health of the mother as well as the baby to find possible complications during pregnancy and delivery. It may also be useful to predict and avoid birth defects since there have been many instances where stress related complications have been known leading to cognitive disorders in the child. The goal of this study is to design and develop a prediction model for stress-based complications during pregnancy, based on the Physical, Social, Environmental and Biological factors. For this the dataset was generated using personalized interview-based survey administered to women who have undergone pregnancy and delivery in the past. The questions were based on the factors mentioned above. The data generated is used to check the correctness of the hypothesis and to evaluate the performance of the proposed stress prediction model using different machine learning algorithms like Support Vector Machine (SVM), Naive Bayes (NB), K-Nearest Neighbor (KNN) and Decision Tree (DT). The experimental results proved that the proposed model achieved an accuracy of 90% when Naive Bayes algorithm was used. The other algorithms produced lesser results but stillclose.

I. Introduction

Stress now has become a vital part of everyone's life. Stress is defined as a sentiment of being under lot of mental or emotional pressure. People have different ways of reacting to stress, a situation that feels stressful to one person may, in fact, be motivating to another. Stress can be explained as by our body method for reacting to any sort of interest. At the point when individuals feel worried by something going ahead around them, their bodies respond by discharging synthetic substances into the blood.