MSTL - AN OVERVIEW

It does this by evaluating the prediction errors of the two types around a particular interval. The take a look at checks the null speculation that the two versions provide the exact same performance on typical, in opposition to the alternative that they don't. In the event the test statistic exceeds a crucial price, we reject the null hypothesis,