Linear Convergence of Stochastic Block-Coordinate Fixed Point Algorithms
Linear Convergence of Stochastic Block-Coordinate Fixed Point Algorithms
复制标题
DOI：
10.23919/eusipco.2018.8552941
复制
发表时间：
2018-09
期刊：
影响因子：
--
通讯作者：
P. Combettes;J. Pesquet
中科院分区：
文献类型：
--
作者：
P. Combettes;J. Pesquet
最近的随机块坐标定点算法特别适合于信号和图像处理中的大规模优化。这些算法的特点是随机扫描规则，以任意选择在迭代过程中激活的变量块，并且它们允许在运算符求值中存在随机误差。本文给出了新的线性收敛结果。在理论和实验上，将这些收敛速度与标准确定性算法在图像恢复问题中的收敛速度进行了比较。
Recent random block-coordinate fixed point algorithms are particularly well suited to large-scale optimization in signal and image processing. These algorithms feature random sweeping rules to select arbitrarily the blocks of variables that are activated over the course of the iterations and they allow for stochastic errors in the evaluation of the operators. The present paper provides new linear convergence results. These convergence rates are compared to those of standard deterministic algorithms both theoretically and experimentally in an image recovery problem.