The Generalized STAR Modeling with Heteroscedastic Effects
In general, the Generalized Space Time Autoregressive (GSTAR) model of space-time assumes constant error variance. In this study, a GSTAR model was built with an error variance that was not constant or had a heteroscedasticity effect, namely the combination of GSTAR–Autoregressive Conditional Hetero...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Mathematics Department UIN Maulana Malik Ibrahim Malang
2022-03-01
|
| Series: | Cauchy: Jurnal Matematika Murni dan Aplikasi |
| Subjects: | |
| Online Access: | https://ejournal.uin-malang.ac.id/index.php/Math/article/view/13097 |
| Summary: | In general, the Generalized Space Time Autoregressive (GSTAR) model of space-time assumes constant error variance. In this study, a GSTAR model was built with an error variance that was not constant or had a heteroscedasticity effect, namely the combination of GSTAR–Autoregressive Conditional Heteroskedasticity (ARCH). The parameters of the GSTAR–ARCH model were estimated using the Generalized Least Square (GLS) method to obtain an efficient parameter estimation. As a case study, the GSTAR–ARCH model was applied to the daily mean wind speed data of New Orleans, Florida and Mississippi to predict the occurrence of Hurricane Katrina that occurred in 2005. The results obtained show that the GSTAR model (3;0,0,1)–ARCH(1) predicts Hurricane Katrina very well. |
|---|---|
| ISSN: | 2086-0382 2477-3344 |