I am working on machine learning from the basics. I want to do it the right way and I am learning probability and statistics. At school and college, mathematics was my favorite subject and I loved treating every problem as a new challenge leading to a fresh perspective. Then life intervened. Now, after a few years, I realize, much to my chagrin, that most of the higher level math I had learned, has evaporated. So there is much I have to relearn. Also, unfortunately, unlike college, it is not possible for me to sit x hours everyday to learn. Currently, I am learning in bursts, sometimes followed by long passive time. So I often have to revise and reorient myself when I sit to study again. I would like to avoid this. Has anyone faced this? How do you tackle it?One solution seems to be using Spaced Repetition cards, like Anki. But it to seems to have it's limitations. http://lesswrong.com/lw/juq/a_vote_against_spaced_repetition/
Obviously, whatever I learn in math, I try to apply it to a lots of problems. But inevitably, there is some fading of knowledge as time passes, and I might not be able to keep redoing the problems.
Please share your experiences. Thank you.