Uplift modelling, also known as incremental modelling is a technique that directly models the incremental impact of a treatment (such as a direct marketing action) .


Uplift  Formula =  (selection % / everything %) -  1

Example: (6.5/3.49)- 1 = 0.8602 --> 86.02%

As we can see, in this case Circolo has an Uplift Score of 86.02% which means that Circolo appears, in the selection that we have done, 86.02% more times than in the original distribution (whole dataset).

This allows us to identify relevant information from the selection that we have done and discover unexpected insights from the selection.

Did this answer your question?