Linear algebra fits clearly into the field of mathematics and doesn't have anything particular to do with AI except that linear algebra might be used as part of an approach, but only to the extend that set theory or vector multiplication might.
Linear regression is a Statistics 101 curve fitting method, with simple formulae for slope, intercept, and correlation coefficient. A basic 1980 pocket calculator has a button for it.
Neither of these tags are particularly AI centric.
They're both core topics that are important to understand well, very important basics, before people can move on to a plethora of more advanced topics in AI. So yes, they absolutely should be tags in AI.se.
It depends on how the questions are asked. Linear Regression is the foundation for many different machine learning and artificial intelligence algorithms. If someone were to ask a question on how their problem could be formatted as a regression, then I would argue that it's perfectly relevant to this SE. Technically, linear regression alone is one of the simplest forms of machine learning. Now, if you were to ask to prove the bounding condtions of certain types of optimizations under purely theoretical conditions, it may not be as relevant to this SE as cross validated for example.
In short, it depends on how the question is asked. If it deals more with the application side of AI, then yes. I think it is perfectly reasonable to ask on the AI SE.
