Prediksi Laju Inflasi dengan Metode Long Short-Term Memory (LSTM) Berdasarkan Data Laju Inflasi dan Pengeluaran Kota Ternate

Abstrak

Inflation is one of the main indicators that reflect the economic stability of a region. Ternate City, as one of the cities in North Maluku Province, exhibits fluctuating inflation dynamics from year to year. This study aims to forecast the inflation rate in Ternate using the Long Short-Term Memory (LSTM) method, which is a neural network architecture well-suited for processing time series data. The data used consists of monthly Consumer Price Index (CPI) figures for Ternate from 2016 to 2023, obtained from the Central Bureau of Statistics (BPS). The LSTM model was trained using monthly CPI changes as the basis for calculating inflation. The model evaluation results show a Root Mean Square Error (RMSE) of 0.9275, Mean Absolute Error (MAE) of 0.8369, and Mean Absolute Percentage Error (MAPE) of 20.13%. These results indicate that the LSTM model performs well in forecasting inflation in Ternate City and can be utilized as a decision-support tool in regional economic planning and policymaking.