Optimized Model for Predicting Gestational Diabetes using ML Techniques

Gestational Diabetes Mellitus (GDM) is a type of diabetes that occurs in women during their gestation period. Some women develop higher blood glucose levels during their gestation. If GDM is ignored or untreated then it can result in permanent medical problems for the baby in the future. This is identified as a serious open problem that is used for good prediction purposes of how to detect the GDM at the starting stage of gestation. For this ML models like Artificial Neural Network, Feed Forward Neural Network, Multi-Layer Perceptron, Naïve Bayes, Gradient Boosting are applied. Among all the Models Gradient Boosting got an accuracy of 0.77. Based on the Gradient boosting model it predicts the framework GDM. This model helps us to find whether the patient suffers from diabetes or not at an early stage to prevent it from further.

I. Introduction

GDM is a kind of diabetes that is diagnosed for the first time during the period of gestation [1]. During this period, maternal insulin resistance is increased to supply sufficient nutrients for the development of the foetus. It develops Hyperglycemia if there isn't enough insulin and enough secretion also. The study of Pregnancy Outcome and hyperglycemia is to explore the large multiethnic cohort issue [2]. The most crucial key in the medical care for a gestational diabetes sufferer is by a controlled blood glucose level is nutrition management [3]. The purpose of identifying whether the patient is suffering from diabetes or not is to avoid the further risk of perinatal outcomes.