http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Monte Carlo simulation of error assumptions in generalized star (1;1) model
DEBBY MASTERIANA,UTRIWENI MUKHAIYAR 장전수학회 2019 Proceedings of the Jangjeon mathematical society Vol.22 No.1
This study is aimed to compare several denitions of error assumption in Generalized Space Time Autoregressive (GSTAR) (1;1) model. This denition is needed since the assumption of normal and identically distributed (iid ) is dicult to be satised. Four assump- tions are simulated in this article respectively spatial correlated error, time correlated error, spatial and time correlated error with martingale dierence, and spatial and time correlated error with beta estimator. By using Monte Carlo simulation, a thousand replications in thirteen locations are done towards GSTAR (1;1) model with varied amount of random data. The result shows that convergence of generalized least square (GLS) estimation approaches the real parameter is faster gained by spatial and time correlated error with martingale dierence assump- tion. This new denition of error assumption provides a more precise GSTAR (1;1) model.