Depends what level you are comfortable with. Given that the kalman filter makes certain statistical assumptions (normally distributed) you need to be familiar w

Depends what level you are comfortable with. Given that the kalman filter makes certain statistical assumptions (normally distributed) you need to be familiar with at least statistics, and know what mean/covariance matrices are. If so you should be able to follow the derivation in https://sites.ualberta.ca/~dwiens/stat679/meinhold&singpurwalla.pdf https://sites.ualberta.ca/~dwiens/stat679/meinhold&singpurwa... up until section 4. (Beyond section 4 it's really just a matter of grinding out the math, and while necessary for implementing it in a practical fashion doesn't really give you any additional understanding).