Risk Estimators for Choosing Regularization Parameters in IllPosed Problems  Properties and Limitations
Abstract
This paper discusses the properties of certain risk estimators recently proposed to choose regularization parameters in illposed problems. A simple approach is Stein's unbiased risk estimator (SURE), which estimates the risk in the data space, while a recent modification (GSURE) estimates the risk in the space of the unknown variable. It seems intuitive that the latter is more appropriate for illposed problems, since the properties in the data space do not tell much about the quality of the reconstruction. We provide theoretical studies of both estimators for linear Tikhonov regularization in a finite dimensional setting and estimate the quality of the risk estimators, which also leads to asymptotic convergence results as the dimension of the problem tends to infinity. Unlike previous papers, who studied image processing problems with a very low degree of illposedness, we are interested in the behavior of the risk estimators for increasing illposedness. Interestingly, our theoretical results indicate that the quality of the GSURE risk can deteriorate asymptotically for illposed problems, which is confirmed by a detailed numerical study. The latter shows that in many cases the GSURE estimator leads to extremely small regularization parameters, which obviously cannot stabilize the reconstruction. Similar but less severe issues with respect to robustness also appear for the SURE estimator, which in comparison to the rather conservative discrepancy principle leads to the conclusion that regularization parameter choice based on unbiased risk estimation is not a reliable procedure for illposed problems. A similar numerical study for sparsity regularization demonstrates that the same issue appears in nonlinear variational regularization approaches.
 Publication:

arXiv eprints
 Pub Date:
 January 2017
 DOI:
 10.48550/arXiv.1701.04970
 arXiv:
 arXiv:1701.04970
 Bibcode:
 2017arXiv170104970L
 Keywords:

 Mathematics  Statistics Theory;
 Mathematics  Numerical Analysis;
 Mathematics  Optimization and Control