Excluding Multiplying Trends Is a Good Idea

We have eight exponentials, moving models in our reduced set. And this are a model that contains only level, no trend or tio nint,. A model that contains strength, a model with sitionality, and a model with strength and susionality. And then we have another four models for a multiplicative type of errow. So we have four models for aritive type of error and four models for the multiplicae type ferrow. Over all, models that we exclude are models with a multiplicative strength, which have shown to have a se forecasts.