Sampled limited memory methods for massive linear inverse problems

Julianne Chung,Matthias Chung,J Tanner Slagel,Luis Tenorio
DOI: https://doi.org/10.1088/1361-6420/ab77da
IF: 2.408
2020-04-30
Inverse Problems
Abstract:In many modern imaging applications the desire to reconstruct high resolution images, coupled withthe abundance of data from acquisition using ultra-fast detectors, have led to new challenges inimage reconstruction. A main challenge is that the resulting linear inverse problems are massive.The size of the forward model matrix exceeds the storage capabilities of computer memory, or theobservational dataset is enormous and not available all at once. Row-action methods that iterateover samples of rows can be used to approximate the solution while avoiding memory and dataavailability constraints. However, their overall convergence can be slow. In this paper, weintroduce a sampled limited memory row-action method for linear least squares problems, where anapproximation of the global curvature of the underlying least squares problem is used to speed upthe initial convergence and to improve the accuracy of iterates. We show that this limited memorymethod is a generaliz...
mathematics, applied,physics, mathematical
What problem does this paper attempt to address?