ESTIMASI PARAMETER DAN PENGUJIAN HIPOTESIS MODEL GEOGRAPHICALLY WEIGHTED GENERALIZED GAMMA REGRESSION

Hasbi Yasin (Departemen Statistika, Fakultas Sains dan Matematika, Universitas Diponegoro Departemen Statistika, FSAD, Institut Teknologi Sepuluh Nopember)
Purhadi Purhadi (Departemen Statistika, FSAD, Institut Teknologi Sepuluh Nopember)
Achmad Choiruddin (Departemen Statistika, FSAD, Institut Teknologi Sepuluh Nopember)

Jurnal Gaussian
Vol 11, No 1 (2022): Jurnal Gaussian

Each location has unique characteristics, which are different from other locations which give rise to spatial effects between locations. Therefore, the Generalized Gamma Regression (GGR) model is not suitable to be applied to this problem. The solution is to use a Geographically Weighted Generalized Gamma Regression (GWGGR) model which produces different parameters for each observation location. This study aims to estimate GWGGR parameters using the Berndt-Hall-Hall-Hausman (BHHH) algorithm. After parameter estimation is performed, the hypothesis testing procedure is used to test the similarity of parameters between the generalized gamma regression and GWGGR and to test the significance of the independent variables in the model, either simultaneously using the Maximum Likelihood Ratio Test (MLRT) or partially using the Z-test. Keywords: BHHH, Generalized Gamma, GGR, GWGGR, MLRT.