Syntha Lab RU

Naive forecast baseline

A simple forecast rule without training, for example “as in the last period”; a complex method or an AI is compared with it.

Forecast and data

Open interactively →

Formula

Improvement = 1 − the method’s error ÷ the baseline error

How to read

A method that does not beat the baseline is not needed however it looks.

Where it is discussed

Topics

Calculations

Documents

Other terms in the group

Open interactively →