MODEL HIBRIDA DEKOMPOSISI-ARIMA UNTUK PERAMALAN INFLASI DI KOTA MAKASSAR

Muhammad Fahmuddin (Program Studi Statistika, Fakultas Matematika dan Ilmu Pengetahuan Alam, Universitas Negeri Makassar, Indonesia)
Zulkifli Rais (Program Studi Statistika, Fakultas Matematika dan Ilmu Pengetahuan Alam, Universitas Negeri Makassar, Indonesia)

Variansi : Journal of Statistics and Its Application on Teaching and Research
Vol 3, No 2 (2021)

Forecasting is an art and predicting science about future events. Forecasting could be basic for short-term, mid-term, and long-term planning. The aim of this study is to create a hybrid decomposition model - ARIMA to forecast inflation data in Makassar City. The decomposition method is used for decomposition the inflation data into trend components, seasonal, and random. Furthermore, the decomposition method could be used to forecasting the tren component dan seasonal. Whereas, the ARIMA method was used to forecasting the random component. The result of this study shows ARIMA model used for forecasting the random component is ARIMA (0,0,[3]) with an AIC score of 171,6973Keywords: Decomposition, ARIMA, inflation