SEEMINGLY UNRELATED REGRESSION (SUR) ESTIMATION OF GSTARX(1,1) MODEL  USING R

Soekardi Hadi Prabowo; Lisana Sumarah Pratignyo

Soekardi Hadi Prabowo Program Studi Matematika Universitas Islam As Syafiiyah
Lisana Sumarah Pratignyo Program Studi Matematika Universitas Islam As Syafiiyah

Keywords: OLS, GSTARX(1.1), Contemporaneous, SUR

Abstract

In this paper we studied SUR estimation parameter of GSTARX(1,1) Model, which used overcome the weakness Ordinary Least Square (OLS) ignoring the information Errors between the equations of each location are correlated so that there is a problem Contemporaneous correlation, which willresult in inefficient estimators SUR. We built the script using R software to estimate the parameter using seemingly unrelated regression Method. This case study to apply the GSTARX model to Contaminant Load of BOD With Exogenous Variable of COD in Watershed Citarum River is significance for the result test hypothesis simultaneous model.