Abstract

ABSTRACT


Inflation is one of the economic phenomena that is always interesting to discuss mainly related to its broad impact on macroeconomics, such as economic growth, external balance, competitiveness, interest rates, and even income distribution. Inflation plays an important role in determining economic conditions, so it needs to get serious attention from various circles, especially the monetary authority responsible for controlling inflation. If inflation can be predicted with high accuracy, it can certainly be used as a basis for government policy making in anticipating future economic activity. This study aims to produce inflation forecast data. The method used in this study is the SARIMA method and Single Exponential Smoothing. In forecasting the rate of inflation where data is in the form of time series, the SARIMA method can show more accurate forecasting results than using the SES method. Based on the comparison of the overall forecasting model, the SARIMA model (2,1,1)(1,1,1) 11  has the smallest error value with a MAD value of 0.117, MSE 0.023 and 0.72% for MAPE. From these results it was collected that the inflation forecasting in Bandung using the SARIMA method has a high accuracy value.


Key words: Inflation, Forecasting, SARIMA, SES