Geographically Weighted Regression Analysis on Cases of Malnutrition Under Five in the West Sumatra

Abstract

Malnutrition is a condition experienced by toddlers due to low nutrition or nutritional needs that have not been met. Effort to improve health status by improving the nutritional status of toddlers. The purpose of this research is clarify the Geographically Weighted Regression model is the best model when compared to the multiple linear regression model. The data used was obtained data from the West Sumatra Provincial Helath Office for 2020. The dependent variable used is the percentage of cases of malnutrition and there are several independent variables, namely the percentage of children under five who were given vitamin A, the percentage of babies who were exclusively breastfeed, and babies born with low birth weight. The results showed that the gwr model can explain the diversity of cases of malnutrition of children under five by 99% with a total squared error of 0.002 compared with multiple linear regression model which is able to explain the diversity of cases of malnutrition among children under five by 43% with a total squared errors of 0.585. It is concluded that the gwr model is the best model.