Generalized Space Time Autoregressive Integrated Moving Average (GSTARIMA) Model to Forecast Cocoa Export Volume

Abstract

Generalized Space Time Autoregressive (GSTAR) is one of time series model used to forecast the data
consisting the element of space and time. This model is limited to the stationary and non-seasonal data. Generalized
Space Time Autoregressive Integrated Moving Average (GSTARIMA) is GSTAR development model that
accommodates the non-stationary and seasonal data. In this research, the model was applied to the monthly cocoa export
volume data from DKI Jakarta, Jawa Tengah and Jawa Timur in the last 8 years. Indonesian cocoa export volume in the
third position in the world trade, after Ivory Coast and Ghana. Identification of the AR and MA are using the minimum
value of AIC. Spatial order is chosen in first order because all of the provinces in this research are located in one island.
From the two spatial weight matrix, which distance inverse and normalized cross-correlation between locations to the
corresponding lag, we have the minimum MSE value to the data is distance inverse.
Keywords: Space time, GSTAR, GSTARIMA, Cocoa