An Implementable First-Order Primal-Dual Algorithm for Structured Convex Optimization

Figure 1

(a) Relative errors of low-rank matrix 𝐿. (b) Relative errors of sparse matrix 𝑆.

(a)

(b)