Estimasi Model Regresi Panel Poisson dengan Conditional Maximum Likelihood
Abstrct: Panel Poisson Regression Model using
Conditional Maximum Likelihood is a combination of cross-section data and time
series data, that is applied to data that individual effects have highly
significant correlation to the predictor variable from a large population, stated
as the following below:.
The purpose of
this final project is to
obtain estimates of
the panel poisson regression model using the Conditional
Maximum Likelihood method and to test the suitability of the model. To estimate
the model parameters
can be obtained
by solving the
equation form below
Parameter estimation of the panel Poisson regression model is gotten in a
implicit form, so that it is solved using numerical iteration, which is the
Newton-Raphson algorithm. After
obtaining the parameter
estimates, carried out
several tests: to
test the parameter estimation twice: simultaneously
using Likelihood Ratio Test (LRT) and individually using test statistics .
After that, continued to test the suitability of the model using deviance test statistic.
Keywords: Conditional Maximum
Likelihood, Deviance, Newton-Raphson, Panel Poisson Regression Model
Penulis: Friska Panggabean,
Suliyanto & Toha Saifudin
Kode Jurnal: jpmatematikadd130072
