PEMODELAN RISIKO PENYAKIT PNEUMONIA PADA BALITA DI PROVINSI JAWA TIMUR DENGAN PENDEKATAN GEOGRAPHICALLY WEIGHTED LOGISTIC REGRESSION

Abstract

This research is aim to determine the comparison of logistic regression models and models Geographically Weighted Logistic Regression and the factors that significantly affect the risk of pneumonia in toddlers in East Java Province. Logistic regression is a statistical analysis that is used to describe the response variable is categorical with the independent variables are categorical or continuous. The main problem of this method if  it’s applied in data that is affected of geographic location or spatial data. One of many method to solve the spatial data is Geographically Weighted Logistic Regression (GWLR). GWLR is a statistical method for analyze the data to account for spatial factor. The results showed that there are no significant differences between the logistic regression model with GWLR model. Factors that significantly affect the risk of pneumonia in toddlers in East Java Province is the percentage of low birth weight, the percentage of  toddlers who get measles immunization, the percentage of toddlers who get vitamin A, and the percentage of toddlers who get DPT+HB immunization.