Abstract Maternal health is the foremost characteristic of women’s health during pregnancy. Particularly, during pregnancy, several health aspects like age, heart rate, blood disorders, etc. can give rise to pregnancy complexities. This designing a Deep Neural Network based York Regression with Swish Hyperbolic Tangent (DNN-YRSHT) method for Pregnancy Health Risk classification. The proposed method for pregnancy patient risk classification comprises of five layers as, input are passed to input layer that is then transmitted to first hidden layer. In first hidden layer Hybrid Class Balanced Standardization-based Pre-processing is applied to class balancing and normalization process separately. In the second hidden layer, York Regression-based Feature Selection model is using relevant further processing. Third hidden layer, Swish Hyperbolic Tangent Activation Function is applied to classify health risk level into two classes, namely low risk or high risk and is transmitted to output layer. In this manner, maternal risk level classification is performed with higher accuracy and computationally efficient manner. The DNN-YRSHT method can assist medical personnel in making quick decisions, improving level of care provided to expectant mothers and their unborn children in a timely manner. Keywords: Deep Neural Network, Hybrid Class Balanced Standardization, York Regression, Swish Hyperbolic, Tangent Activation Function
MidhunaMurali et al. (Sun,) studied this question.