Mixed Geographically Weighted Poisson Regression Model in The Number of Maternal Mortality.

Rendra Erdkhadifa, 199002102019031014 (2022) Mixed Geographically Weighted Poisson Regression Model in The Number of Maternal Mortality. Jurnal Mantik, 6 (3). pp. 3512-3521. ISSN 2685-4236

Text Mixed Geographically Weighted Poisson Regression Model in The Number of Maternal Mortality.pdf Download (739kB)

Abstract

This study aims to prove and to use the multiple regression analysis methods can be developed into Poisson regression because the total data follows the assumption of a Poisson distribution. Conditions that occur in poisson regression obtained a global regression coefficient value, which means that each observation point has generalized observation characteristics influenced by the same variables. Based on this problem, a spatial regression is developed where the geographical weighting that called geographically weighted Poisson regression. The results obtained in the geographically weighted poisson regression method allow for variables that have an effect at all observation points. So developed by geographically weighted Poisson regression semiparametric method. This method is implemented in solving the problem of maternal mortality in East Java which is thought to be influenced by the some independent variables. In this study, using East Java data in 2019. The results of mixed geographically weighted poisson regression results, with kernel function used in the analysis is fixed gaussian and the variable number of health facilities was used as a global variable, and obtained 12 groups with local variables that both had a significant effect.

Actions (login required)

View Item