A- LINEARIZED BREGMAN ITERATION ALGORITHM
A-线性Bregman 迭代算法

Linearized Bregman iteration is an efficient algorithm in many areas such as compressed sensing which was proposed by Osher and Cai,et al,recently. In this paper, this iteration for the following minimization problem is studied as the matrix A is not surjective: A simplify proof of the convergence of the linear Bregman iteration is given. A novel A- linear Bregman iteration is proposed and its convergence is proved for special case. Numerical results demonstrate that this novel iteration can recovery sparse signal from linear measurements.