作者: Li Zhang , Wei-Da Zhou , Gui-Rong Chen , Ya-Ping Lu , Fan-Zhang Li
DOI: 10.1016/J.KNOSYS.2013.09.007
关键词:
摘要: This paper proposes a decomposition algorithm for sparse signal reconstruction.A small quadratic programming problem is solved in each iteration.The convergence of the also shown this paper.The method can get fast regularization values. In compressed sensing, reconstruction required stage. To find solutions problems, many methods have been proposed. It time-consuming some when parameter takes value. reconstruction, which almost insensitive to parameter. iteration, subproblem or our algorithm. If extended solution current iteration satisfies optimality conditions, an optimal found. On contrary, new working set must be selected constructing next subproblem. The paper. Experimental results show that able achieve