Forecasting Indonesian inflation using a hybrid ARIMA-ANFIS

This paper discusses the prediction of inflation rate in Indonesia. The data used this research is assumed to have both linear and non-linear components. ARIMA model selected accommodate component, while ANFIS method accounts for component data. Thus, known as hybrid ARIMA-ANFIS model. clustering performed using Fuzzy C-Mean (FMS) with a Gaussian membership function. Consider 2 6 clusters. optimal number clusters assessed according minimum value error prediction. To evaluate performance fitted model, it can be compared classical ordinary result reveals that best Indonesia ARIMA(2,1,0). In ARIMA(2,1,0)-ANFIS two are optimal. Meanwhile, optimum six. comparison accuracy confirms superior individual alone either or