<![CDATA[The existence of spatial dependency between regions in meeting food and non-food needs causes the consumer price index (CPI) to be interrelated between one region and others. This research aims to compare the performance of space time autoregressive (STAR) and generalized space time autoregressive (GSTAR) in predicting CPI in the cities of Padang, Pekanbaru, Jambi and Palembang. Moran’s Index testing shows the effect of spatial dependence on the CPI series between the regions of Padang, Pekanbaru, Jambi and Palembang during the period January 2015 – April 2020. The average CPI of Palembang is the smallest and the slowest increase between months compared to the other three cities. The highest average CPI in Padang city. The average CPI in Pekanbaru city is the fastest increase between months. The result of model identification shows that the estimator of the GSTAR(1;1)-I(1) model using a uniform weighting is the best and suitable in estimating CPI of the four cities. The smallest RMSE value indicates CPI forecasting in Padang, Pekanbaru and Palembang is better to use the GSTAR(1;1)-I(1) model with the uniform weighting location. Meanwhile, the Jambi City CPI forecasting is better to use the STAR(1;1)-I(1) model with the uniform weighting location.]]>
Copyrights © 2020
