Each sub-discipline has it's own "state of the art" and the numerical optimization is so diverse now that the question is very difficult to answer. I would look for a recent textbook of monograph and see what sub-discipline really interests me the most and start from there.
You could also look for the techniques use in on application domain you are interested in and look for recent development in this particular technique.
Thank You for Your responses, and sorry for not replying in a timely manner. :) The algorithms that I'm interested in are those for optimizing numerical functions such as Sphere, Rosenbrock, etc.