How much math do you recommend to learn before starting learning machine learning?

I'm currently a student in computational linguistics, where we have scratched the surface of machine learning, learning how to implement and apply simple neural networks and decision trees. I find the subject very fascinating, but at the same time feel like I'm lacking the necessary math skills for actually understanding what's going on. I have some math background already, having finished courses in single-variable calculus and basic linear algebra, and it definitely helped to some degree, but not nearly enough.

My questions are, if I continue taking some extra math courses in linear algebra, multi-variable calculus and probability, will it benefit my understanding of machine learning enough for it being worth it, and is there a point when it's worth stopping and start focusing on actual machine learning?