arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 0912.1660math.OC
收藏

稀疏恢复的基于递归的方法

Gradient-based methods for sparse recovery

William Hager, Dzung Phan, Hongchao Zhang

摘要:

分析了SpaSRA算法(通过可分离近似进行稀疏重建)的收敛速度,该算法用于最小化和$f (\m{x}) + ψ(\m{x})$,其中$f$是光滑的,而$ψ$是凸的,但可能是非光滑的。它表明,如果$f$是凸的,那么在迭代$k$的目标函数中的误差,$k$足够大,是有界的$a/(b+k)$为适当的选择$a$和$b$。而且,如果目标函数是强凸的,则收敛是$R$ -线性的.给出了一种基于循环BB迭代和自适应线搜索的改进算法。该算法的性能进行了研究,在信号处理和图像重建领域的应用。

英文摘要:

The convergence rate is analyzed for the SpaSRA algorithm (Sparse Reconstruction by Separable Approximation) for minimizing a sum $f (\m{x}) + ψ(\m{x})$ where $f$ is smooth and $ψ$ is convex, but possibly nonsmooth. It is shown that if $f$ is convex, then the error in the objective function at iteration $k$, for $k$ sufficiently large, is bounded by $a/(b+k)$ for suitable choices of $a$ and $b$. Moreover, if the objective function is strongly convex, then the convergence is $R$-linear. An improved version of the algorithm based on a cycle version of the BB iteration and an adaptive line search is given. The performance of the algorithm is investigated using applications in the areas of signal processing and image reconstruction.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑