Current browse context:
math.NA
Change to browse by:
References & Citations
Mathematics > Numerical Analysis
Title: Unbiased Bregman-Risk Estimators: Application to Regularization Parameter Selection in Tomographic Image Reconstruction
(Submitted on 17 Sep 2021 (v1), last revised 18 Nov 2021 (this version, v2))
Abstract: Unbiased estimators are introduced for averaged Bregman divergences which generalize Stein's Unbiased (Predictive) Risk Estimator, and the minimization of these estimators is proposed as a regularization parameter selection method for regularization of inverse problems. Numerical experiments are presented in order to show the performance of the proposed technique. Experimental results indicate a useful occurence of a concentration of measure phenomena and some implications of this hypothesis are analyzed.
Submission history
From: Elias Salomão Helou Neto [view email][v1] Fri, 17 Sep 2021 23:41:37 GMT (7086kb,D)
[v2] Thu, 18 Nov 2021 21:39:28 GMT (8349kb,D)
Link back to: arXiv, form interface, contact.