Journal of Statistical and Econometric Methods

Seasonal autoregressive integrated moving average vector models and their application to quarterly rainfall series

• [ Download ]
• Times downloaded: 10207

• Abstract

  In this study, n-dimensional Seasonal autoregressive integrated moving average vector (SARIMAV) models of additive form are compared with univariate models. Ordinary least squares method is adopted to estimate parameters of the models. The bivariate models obtained are reliable as much as univariate seasonal models. It is established that seasonal time series model is not only applicable in univariate case. Hence, seasonal autoregressive integrated moving average vector models are established, verified valid and useful in modeling multiple seasonal series.