Implementation of Time Series Forecasting for Inflation Prediction in Indonesia

Abstrak

Inflation is a crucial macroeconomic indicator that reflects economic stability and influences sectors such as consumption, investment, and policy-making. This study aims to implement and compare three time series forecasting models: Seasonal Autoregressive Integrated Moving Average (SARIMA), Support Vector Regression (SVR), and Long Short-Term Memory (LSTM) to predict inflation in Indonesia. The study utilizes monthly inflation data from Bank Indonesia (2003–2024) and Consumer Price Index (CPI) data from Statistics Indonesia (2003–2019). Model performance is evaluated using Mean Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Squared Error (RMSE). The results indicate that SVR achieves the best performance in predicting inflation, with MAE of 1.53, MSE of 2.72, and RMSE of 1.64, demonstrating its effectiveness in capturing nonlinear patterns. Meanwhile, SARIMA provides the most accurate predictions for CPI, with MAE of 11.55, MSE of 191.76, and RMSE of 13.84. LSTM shows competitive performance but is less consistent compared to the other models. These findings highlight the importance of selecting appropriate models based on data characteristics to improve forecasting accuracy and support economic decision-making.