I'd suggest making machine-learning a synonym of learning-theory.

Currently (Dec 6, 2021)

• machine-learning has 39 watchers, 122 questions. It has no tag info.
• learning-theory has 19 watchers, 70 questions. Its tag info is: "This tag is used for questions that are related with following branches: Statistical learning theory, Machine learning, Vapnik–Chervonenkis theory (VC theory) and all other branches that are studied and applied in the area of learning theory that involves various kinds of mathematics."

I think there is no sensible difference justifying the distinction between these two tags. As regards which one should be made synonym: while the first might look slightly more specific, the fact that it has more occurrences simply reflects that people asking questions in this area rather have in mind the keywords "machine learning". I tend to prefer the second one (i.e., making the first one synonym of the second one, rather than vice versa) because learning-theory is slightly more generic, and makes it clearer that this is a mathematical subject.