Deleted:Regression trees against background

Question

[druggable]

Hi Everyone,

Do you know if there is an algorithm for implementing regression trees (random forest) while taking into account a background/control data? Much like Leo Breiman's unsupervised random forest but for regression? There the null background is a permuted matrix generated from the original matrix. How can I adapt this to regression? Any help would be very much appreciated.

Thanks a lot.