Is dimensionality reduction considered to be primarily unsupervised learning?
I have always thought of dimensionality reduction as unsupervised learning. It is what you usually see in textbooks. "Now we will present several unsupervised learning types: clustering, dimensionality reduction, etc.". But now I saw that there is also supervised dimensionality reduction. It does make sense though, I can imagine supervised dimensionality reduction scenarios.
But, then I wonder why most of the time people refer to dimensionality reduction as unsupervised learning?