Understanding and Enhancing Data Recovery Algorithms
Lists
Only you see this
Abstract
We prove new results about the robustness of noise-blind decoders for the problem of reconstructing a sparse vector from underdetermined linear measurements. Our results imply provable robustness of these decoders for random measurements with heavy-tailed distributions. We further propose a new algorithm for the reconstruction of low-rank matrices from few linear observations or from missing data. The algorithm is based on the iterative minimization of well-designed quadratic models of a non-convex objective and combines data efficiency with scalability and a superlinear local convergence rate. This framework is extended to the completion of structured low-rank matrices.
BibTeX
@inproceedings{kummerle2020understanding,
author = {Christian Kümmerle},
title = {{Understanding and Enhancing Data Recovery Algorithms}},
booktitle = {mediaTUM},
year = {2020},
}Loading related content...