I was initially thinking the same approach, but I cannot figure out how that could determine the relevance of the features. Aren’t we suppose to use single dimensional RVs for ℓ and η, assigning a single RV for all the features?
I was initially thinking the same approach, but I cannot figure out how that could determine the relevance of the features. Aren’t we suppose to use single dimensional RVs for ℓ and η, assigning a single RV for all the features?