I recently finished my CS degree, And want to specialize in the "AI" field, the typical academic route (Master degree -> PHD -> More research) is my goal. But I feel like my mathematical background is not that "mature" yet (I'm aware that learning will never stops).
Thus, I'm looking for a coherent (no circular dependencies as I faced before) mathematical curriculum, perhaps a bottom-up approach? (for example : start with X then Y, Then choose either W or Z,... that kind of stuff)
I don't want to be scratching the surface again, nor switching from a resource to another. Up untilthis date, all my queries were inconclusive. All what I found is some "goal-oriented" stuff (usually something like do linear algebra, or calculus, or probability and statistics, etc) which has no logical form | sequence.
Is it a mixture of information overload and little guidance which causes this? Or is it my initial approach of wanting to learn mathematics in a "low" level way ?
For AI, you must study probability and statistics, information theory, and linear algebra. Once you get into a specific domain (speech & language, vision, planning, etc.), then the required mathematics will become obvious from the papers you read and the guidance from your dissertation advisor.