Table 1. Comparison of three potential ARIMA models.
Model | R2 | MSE | SBC | AIC |
---|---|---|---|---|
ARIMA (0, 1, 1) × (1, 1, 0)12 | 0.7810 | 0.6616 | -25.4473 | -30.2850 |
ARIMA (0, 1, 1) × (0, 1, 1)12 | 0.7790 | 0.6691 | -24.5173 | -29.3550 |
ARIMA (2, 1, 0) × (0, 1, 1)12 | 0.7840 | 0.6521 | -22.2328 | -29.4893 |
ARIMA=the autoregressive integrated moving average; R2 =coefficient of determination; MSE=mean square error; AIC=Akaike information criterion; SBC=Schwarz Bayesian information criterion.