EE is pretty broad, so which courses are useful depends upon your graduate specialization. Having said that, in my opinion all undergrad EE majors really should take linear algebra. Any school that does not force students to take it is simply not doing right by the students. If you didn't take it in undergrad then you must take it during grad school. It is just too fundamental.
After that, it can depend. For the "applied physics" sorts of specializations (electromagnetics, solid state, lasers, ...) partial differential equations and complex analysis are very useful. I knew some who took the "math methods" courses from physics departments. Many of the folks I know who specialized in "applied math" specializations (image processing, communications, statistical signal processing, ...) often took undergrad level real analysis courses, modern algebra, etc., and some took advanced probability, statistics, and/or stochastic processes from either math or Operations Research departments.
good luck,
jason