There is also Reinforcement learning that are not supervised. That is when output is expected to have a direction. Thank you for your interest. But you are correct, Most of Reinforcement Learning is always Unsupervised.
>Interesting. How does it perform out-of-sample? - Soon enough.
One more very strange idea, the system of education causes sometimes limitations to the mind and imagination, by defining terms only to the limits that are ruled by the context of the education system, which also causes the development of ideas to be constrained and limited because of these enforced limitations of definitions, and the lack of encouragement to go beyond known definitions. Therefor I would suggest that Reinforcement Learning can be also supervised. Not only Unsupervised by default.

While in this case, some sort of supervision is needed.