Inflation forecasting for Samarinda city using hybrid singular spectrum analysis-neural network model

A hybrid method is a combination of two or more forecasting methods. One of the hybrid methods is Singular Spectrum Analysis (SSA)-Neural Network (NN). In time series data, SSA performs well in identifying trend, seasonal, cyclical, and noise patterns. NN can process enormous amounts of data and provide high accuracy in predicting. The SSA-NN hybrid model is a good combination for forecasting economic data that tends to have trend and seasonal patterns, such as inflation data. The purpose of this study was to obtain a comparison of the forecasting performance between the SSA model and the best hybrid SSA-NN model and to obtain the results of inflation forecasting for Samarinda City in 2022. Based on SSA modeling in Samarinda City inflation, the Root Mean Square Error (RMSE) value was 0.0435, while the best SSA-NN 3 neuron hybrid model value of RMSE was 0.0160. The best SSA-NN hybrid model performed better than the SSA model because it produced a smaller RMSE value. Inflation forecasting for Samarinda City in 2022 using the hybrid SSA-NN 3-neuron model revealed that the highest inflation was 0.83% in July and the lowest was -0.39% in August.