Classes in Maximum likelihood classification

Hello everybody!
I am trying to perform a supervised classification of the type maximum likelihood classification of an image obtained from Sentinel-1.
I have analyzed how various types of vegetation have changed over the years. For this I have made several data vector polygons for each type of vegetation. What I would like to do is tell the classifier that Class 1 is the one represented by the training vectors Pol_1, Pol_2 and Pol_3; that class 2 is the one represented by the training vectors Pal_1, Pal_2 and Pal_3.
Would anyone know how to do that?

Thank you very much in advance

Did you digitize the polygons in SNAP or outside?

You can digitize multiple polygons belonging to one class as described in this tutorial: Landcover classification with Sentinel-1 GRD (pages 21-22)

If you collected them outside SNAP, you can use a dissolve tool to merge them into one multi-polygon per class.

But you can also re-classify the results and assign one pixel value to Pol1-3 and Pal1-3.