An efficient method for adapting step-size parameters of primal-dual hybrid gradient method in application to total variation regularization
An efficient method for adapting step-size parameters of primal-dual hybrid gradient method in application to total variation regularization
复制标题
DOI：
10.1109/apsipa.2017.8282164
复制
发表时间：
2017-12
期刊：
影响因子：
--
通讯作者：
Tatsuya Yokota;H. Hontani
中科院分区：
文献类型：
--
作者：
Tatsuya Yokota;H. Hontani
原对偶混合梯度法是一种非常重要的凸优化方法，在信号处理中有着广泛的应用，如全变分正则化。与单纯的原始方法或对偶方法相比，该方法具有更新过程计算量小、收敛速度快等优点。然而，初始步长和对偶步长参数的选择困难是其关键瓶颈。在本文中，我们提出了一种新的PDHG自适应步长参数选择方法，这是Goldstein等人在2015年提出的技术的修改版本。实验结果表明，该算法的收敛速度有了很大的提高.
Primal-dual hybrid gradient (PDHG) method is a very important technique for convex optimization which has a lot of applications in signal processing such as total variation regularization. It is efficient for low-computational cost of update procedures and relatively faster convergence compared with only primal or dual method. However, the difficulty for selecting a primal and a dual step-size parameters is well-known as its critical bottleneck. In this paper, we propose a new adaptive step- size parameter selection method for PDHG which is a modified version of a technique proposed by Goldstein et al. in 2015. A great improvement of convergence speed was shown in our experiments.