FORECASTING THE CONSUMER PRICE INDEX WITH GENERALIZED SPACE-TIME AUTOREGRESSIVE SEEMINGLY UNRELATED REGRESSION (GSTAR-SUR): COMPROMISE REGION AND TIME

Abstract

Economic success will provide benefits for improving people’s welfare. An important indicator to determine economic success can be seen through inflation by calculating the Consumer Price Index (CPI). CPI is a time series data that is influenced by elements between locations. The GeneralizedSpace-Time Autoregressive (GSTAR) method is a suitable method to be applied to CPI data because it involves elements of time and location (spatiotemporal). The problem is that the GSTAR model cannot detect any correlated residuals. The GSTAR model was developed into the GSTAR-SUR model to estimate parameters with correlated residuals so produce more efficient estimates. The purpose of this study was to determine the best GSTAR-SUR model to predict the CPI of six cities in Central Java, namely Cilacap, Purwokerto, Kudus, Surakarta, Semarang, and Tegal. The data that used is secondary data sourced from BPS Central Java Province. Based on the results of the analysis, the best model formed is the GSTAR-SUR (11)-I(1) model with an RMSE value of 6.213. Forecasting results show that the CPI value for the next 6 months will increase every month for each city