Stochastic Methods for L1 Regularized Loss Minimization

We describe and analyze two stochastic methods for $\ell_1$ regularized loss minimization problems, such as the Lasso. The first method updates the weight of a single feature at each iteration while t