It does this by comparing the prediction errors of the two models about a certain period of time. The test checks the null speculation which the two designs contain the identical performance on regular, against the alternative that they do not. Should the check statistic exceeds a important value, we reject the null hypothesis, indicating that the