Deconstruct: Break down the math you want to know into big problems and concepts. Pick a math-related goal that is Measurable and Time-Bound
Selection: What the 20% of math concepts, that if made really strong, would solve 80% of math problems
Sequencing: What order of material should you study for maximal progress
Stakes: Find some incentive to complete the problem. Some nice view of mathematical terrain, as part of a masters program, applications to another field, a prize, a cookie. Anything that motivates you to actually make progress towards the goal
This approach helped me learn a bunch of high level math like abstract algebra, analysis, linear algebra, etc.