English

Scalable iterative data-adaptive RKHS regularization

Numerical Analysis 2024-01-02 v1 Numerical Analysis

Abstract

We present iDARR, a scalable iterative Data-Adaptive RKHS Regularization method, for solving ill-posed linear inverse problems. The method searches for solutions in subspaces where the true solution can be identified, with the data-adaptive RKHS penalizing the spaces of small singular values. At the core of the method is a new generalized Golub-Kahan bidiagonalization procedure that recursively constructs orthonormal bases for a sequence of RKHS-restricted Krylov subspaces. The method is scalable with a complexity of O(kmn)O(kmn) for mm-by-nn matrices with kk denoting the iteration numbers. Numerical tests on the Fredholm integral equation and 2D image deblurring show that it outperforms the widely used L2L^2 and l2l^2 norms, producing stable accurate solutions consistently converging when the noise level decays.

Keywords

Cite

@article{arxiv.2401.00656,
  title  = {Scalable iterative data-adaptive RKHS regularization},
  author = {Haibo Li and Jinchao Feng and Fei Lu},
  journal= {arXiv preprint arXiv:2401.00656},
  year   = {2024}
}