Key points are not available for this paper at this time.
We describe, analyze, and experiment with a framework for empirical loss minimization with regularization. Our algorithmic framework alternates between two phases. On each iteration we first perfor...
DuchiJohn et al. (Tue,) studied this question.