In: Economics
Outline the Steps for Estimating an ARIMA model for Forecasting. How can you judge that your forcasts are robust?
1.Model identification
2.Parameter estimation
3.Model checking
model identifications:use plots and summarise statistical to identify the trends, seasonality, and autoregression elements to get out an idea of the amount of difference and the size of the lag that had been require.
parameter estimation: use a fitting procedures to find out the coefficien of the regression models.
model checking: use plots and statisticals tests of the residual error to determine the amount and the type of temporal structure not captured by the models.
the process is repeated untill either a desirable level of fitting that is achieved on the in-sample or out-sample observation.