INVERSE GAUSSIAN REGRESSION MODELING AND ITS APPLICATION IN NEONATAL MORTALITY CASES IN INDONESIA

Inverse Gaussian Regression (IGR) is a suitable model for modeling positively skewed response data, which follows the inverse distribution. The IGR was formed from Generalized Linear Models (GLM). This study aims to with applied factors influencing infant mortality cases of provinces in Indonesia. Estimation parameters employed by Maximum Likelihood (MLE) and Fisher scoring methods. Ratio Test (LRT) Wald test were used hypothesis testing significance parameters. Indonesia 2020. data using obtained Ministry Health Republic Central Bureau Statistics. result shows that based on were: percentage pregnant women who received blood-boosting tablets, low birth weight, complete neonatal visits (KN3), toddlers early initiation breastfeeding, are exclusively primary immunization, households access adequate drinking water, appropriate sanitation.