Inflation is a crucial economic indicator, and its growth rate is always aimed to be low and stable to prevent macroeconomic disruptions that might lead to economic instability. Considering the significant impact it can have, predicting future inflation values is essential. This study focuses on forecasting inflation, particularly in the province of West Java, using the ARIMA, Linear Regression, and Triple Exponential Smoothing methods. The goal is to find the method that yields the lowest error to ensure more accurate forecasting results. The research employs inflation data from June 2009 to May 2023 in West Java, collected from the Badan Pusat Statistik (BPS) of West Java Province. The study involves several stages: (1) collecting inflation data, (2) preprocessing the data, (3) constructing forecasting models and obtaining results, and (4) comparing accuracy outcomes. After comparing the methods, it was found that the Triple Exponential Smoothing method emerged as the most effective one. This method exhibited the lowest error evaluation, with an RMSE value of 0.1719703, indicating good accuracy and suitability for forecasting inflation values in the province of West Java for the future
Copyrights © 2025