1
$\begingroup$

I am planning to join a CS Ph.D. program in 6 months. My topics of research will be in the area of Reinforcement Learning and Game Theory. Even though I have a good grasp of these two topics’ applications and empirical side, I struggle with the theory a lot when reading papers. For example, It is taking me a lot of time to understand the regret analysis of advanced Bandits. I have taken most of the essential math courses during my undergrad, including Differential + Integral + Multivariate calculus, Linnear Algebra, Probability, Stat, etc. But I feel most of those courses did not go deep enough. So I want to revisit some of my undergrad math and would also like to learn additional topics. Given my area of Ph.D., I would like to know the essential math topics I should study in the next six months to have a good grasp on the theoretical side and do well in graduate-level courses consistent with such a Ph.D.

$\endgroup$
1
  • $\begingroup$ I don't have a good answer but I would suggest to just deep dive into, say 10, relevant papers. And by deep dive I mean you being able to reconstruct the proofs on paper and really understand it. Nothing will beat that. Any tools that you don't know can be looked upon on a need to know basis. $\endgroup$
    – karmanaut
    Dec 22 '20 at 17:48
4
$\begingroup$

Know linear algebra well, say, at the level of Peter Lax' book (start with the first 9 chapters). Also, some basic real analysis and probability theory should be a good place to start.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.