Considering that random forests have many layers and beat deep learning in most applications, maybe we just need to rebrand them as deep forests and they'll be the next big thing.
-
Putting performance and accuracy aside, how can we formulate things like Language Models, Auto Encoders and Multi-output Classifiers in the Random Forrest Gump world? I believe NNs won mostly for their flexibility, even without their performance numbers
Those have all been done, FYI (e.g., by using them in graphical models).
