<![CDATA[Inflation is one of the main macroeconomic indicators in Indonesia. Inflation occurs when demand exceeds supply, and if not properly controlled, it may affect the economic stability of a region. Inflation forecasting is therefore essential as a basis for governments in formulating and evaluating economic policies. This study aims to compare the performance of the Seasonal Autoregressive Moving Average (SARIMA) method and the Bayesian Structural Time Series (BSTS) method in forecasting inflation in East Nusa Tenggara Province. SARIMA is a classical forecasting method designed to handle seasonal patterns, while BSTS is a state-space model that allows separate decomposition of trend, seasonal, and regression components. The results of this study indicate that the BSTS method outperforms SARIMA, as reflected by smaller forecast error values. The BSTS model with a Semilocal Linear Trend component produces an RMSE of 0.5893397, an MAE of 0.4759239, and a MASE of 0.6509315.]]>
Copyrights © 2026
