Comparison of Fuzzy Time Series Methods and Autoregressive Integrated Moving Average (ARIMA) for Inflation Data

This study compares the Fuzzy Time Series (FTS) method with the Autoregressive Integrated Moving Average (ARIMA) method on time series data. These two methods are often used in predicting future data. Forecasting or time-series data analysis is used to analyze data in the form of time series. In this study, Indonesian inflation data was used to be analyzed in comparing the FTS and ARIMA methods. Inflation is one of the economic indicators used to measure the success of a country's economy. If the inflation rate is low and stable, it will stimulate economic growth. This inflation value is calculated every month where the value can decrease and increase from one period to another. Comparison of the FTS and ARIMA methods is seen in the error value generated by the two methods. A method can be better than other methods if the value of the resulting forecast error is smaller. In this study, Mean Squared Error (MSE) and Mean Absolute Percentage Error (MAPE) were used to see the magnitude of the error value of the two methods on the fives data testing used. The results obtained in this study are the results of Indonesia's inflation forecast for the period January to May 2021 using the FTS method, respectively, at 0.57%, 0.375%, 0.2%, 0.2%, and 0.1125%, while the forecast results using the ARIMA method, respectively. Of 0.3715848%, 0.2362817%, 0.1508295%, 0.1731906%, and 0.2432851% and the results of calculating the size of error using MSE and MAPE indicate that the ARIMA method with the model ARIMA(3,0,0) is better at predicting inflation data in Indonesia with a value of MSE of 0.009 and MAPE of 64.987% compared to the FTS method resulted in MSE values of 0.047 and MAPE of 132.548%.