Select the analysis options for Forecast with Best ARIMA Model

Specify which information criterion to use to select the best ARIMA model.

Both AICc and BIC assess the likelihood of the model and then apply a penalty for adding terms to the model. The penalty reduces the tendency to overfit the model to the sample data. This reduction can yield a model that performs better in general.

As a general guideline, when the number of parameters is small relative to the sample size, BIC has a larger penalty for the addition of each parameter than AICc. In these cases, the model that minimizes BIC tends to be smaller than the model that minimizes AICc.

When the sample size is small relative to the parameters in the model, AICc performs better than AIC. AICc performs better because with relatively small sample sizes, AIC tends to be small for models with too many parameters. Usually, the two statistics give similar results when the sample size is large enough relative to the parameters in the model.

In Confidence level, enter the level of confidence for the probability limits of the forecasts. The probability limits treat the forecast value as a random variable.

Usually, a confidence level of 95% works well. For the probability limit for a forecast value, 95% indicates that the the probability that the forecast value falls into the interval that the limits define is 0.95.

Use a Box-Cox transformation of a time series to try to make the variance of the series stationary. Stationary variance is a requirement for an ARIMA model. Use a time series plot to determine if the variance of a time series is stationary. If the time series has a pattern in the spread of the points, then the variance is not stationary.