Greedy randomized Kaczmarz method for solving system of absolute value equations
摘要
The greedy randomized Kaczmarz method is effective for iteratively solving large-scale linear systems. Building on its promising convergence properties and good numerical behavior, we extend its methodological framework to the more challenging problem of solving the system of absolute value equations. By detailed analysis, we establish an efficient methodology for such problems. Additionally, we obtain the convergence of this method under certain conditions. Finally, we substantiate its efficiency via numerical experiments.