Parameter Estimation STAR (1;1) Model Using Binary Weight

Khafsah Joebaedi (Department of Mathematics, Faculty of Mathematics and Natural Science (FMIPA), Universitas Padjajaran, Bandung, Indonesia)
Iin Irianingsih (Department of Mathematics, Faculty of Mathematics and Natural Science (FMIPA), Universitas Padjajaran, Bandung, Indonesia)
Badrulfalah Badrulfalah (Department of Mathematics, Faculty of Mathematics and Natural Science (FMIPA), Universitas Padjajaran, Bandung, Indonesia)
Dwi Susanti (Department of Mathematics, Faculty of Mathematics and Natural Science (FMIPA), Universitas Padjajaran, Bandung, Indonesia)
Kankan Parmikanti (Department of Mathematics, Faculty of Mathematics and Natural Science (FMIPA), Universitas Padjajaran, Bandung, Indonesia)

Eksakta : Berkala Ilmiah Bidang MIPA
Vol. 20 No. 2 (2019): Eksakta : Berkala Ilmiah Bidang MIPA (E-ISSN : 2549-7464)

Space Time Auto Regressive(1;1) Model or STAR(1;1) model is a form of model that involves location and time. The STAR(1;1) model is a stationary space time model in mean and variance. The STAR model can be used to forecast future observations at these locations by involving the effects of observations at other nearby locations in spatial lag 1 and lag time 1 [2]. The STAR model can be written as a linear model assuming that error is normally distributed with zero mean and constant variance. In this research, the parameter estimation procedure for STAR model using binary weight, MKT method and STAR model for the estimation of petroleum production in 3 wells is assumed to be in a homogeneous reservoir.