1: \begin{abstract}

2: We propose  a  decomposition framework for the parallel optimization

3: of the sum of a differentiable function and a (block) separable

4: nonsmooth, convex one. The latter term is typically used to enforce structure in the solution

5: as, for example, in Lasso problems.

6: Our framework is very flexible and includes both fully parallel Jacobi schemes

7: and Gauss-Seidel (Southwell-type) ones, as well as virtually all possibilities in between (e.g.,  gradient- or Newton-type

8: methods) with only a subset of  variables  updated at each iteration. Our theoretical convergence results improve on existing ones, and

9:   numerical results show  that the new method compares favorably to existing

10: algorithms.

11:  %\vspace{-0.4cm}

12: \end{abstract}

13: