Penerapan Pemodelan Space-Time Pada Data Inflasi Bulanan Di DKI Jakarta, Depok, Dan Bogor Tahun 2017-2021

Abstract

Inflation is a condition where commodity prices experience a continuous increase, and increases in the prices of goods and services that occur tend to expand and result in the prices of other goods. Developments regarding time series analysis led to the idea that data does not only have an attachment to the previous time, but also location. The combination of these two methods is hereinafter referred to as the space-time model. The space time models that are often used are STAR (Space Time Autoregressive) and GSTAR (Generalized Space Time Autoregressive). The STAR model is more suitable for homogeneous locations, while the GSTAR model is used for data with heterogeneous characteristics. In determining the model for monthly inflation rate data for 2017-2021 in DKI Jakarta, Depok and Bogor, the candidate models STARIMA (1.1) and GSTARIMA (1.1) were obtained. Based on the RMSE value of each model, it can be concluded that the STARIMA model (1.1) is better to use because it has a smaller RMSE value than GSTARIMA (1.1).