Possibly you don't want to really ask for "collaboration" as much as "advice" or "feedback". Advice and feedback are "limited" responsibilities, while collaboration is an on-going thing. People might be willing to give advice but not make any long-term commitment.
Also, keep in mind that most professional mathematicians, by their own design, are already "fully-booked" for the foreseeable future, apart from possible large surprises which would justify changing their schedule abruptly.
That is, they already have many on-going projects, as many collaborators as they'd want (if not too many!), as well as other obligations. So, imagine that you are asking a person who is already extremely busy to do something extra. Not that that's impossible, but "it's a big ask". :)
EDIT: ... and, anyway, as in other comments and answers, you don't have to have collaboration of a professor to attempt to publish (in the old sense of "in a refereed journal") your work. Yes, the stylistic expectations, and judgement about level of innovation and so on, will be easier for an experienced person, but in that regard all you really need or want is "advice" rather than "collaboration".