Robust Sparse Estimation for Gaussians with Optimal Error under Huber Contamination
Robust Sparse Estimation for Gaussians with Optimal Error under Huber Contamination
About this item
Full title
Author / Creator
Publisher
Ithaca: Cornell University Library, arXiv.org
Journal title
Language
English
Formats
Publication information
Publisher
Ithaca: Cornell University Library, arXiv.org
Subjects
More information
Scope and Contents
Contents
We study Gaussian sparse estimation tasks in Huber's contamination model with a focus on mean estimation, PCA, and linear regression. For each of these tasks, we give the first sample and computationally efficient robust estimators with optimal error guarantees, within constant factors. All prior efficient algorithms for these tasks incur quantitat...
Alternative Titles
Full title
Robust Sparse Estimation for Gaussians with Optimal Error under Huber Contamination
Authors, Artists and Contributors
Author / Creator
Identifiers
Primary Identifiers
Record Identifier
TN_cdi_proquest_journals_2962943913
Permalink
https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_proquest_journals_2962943913
Other Identifiers
E-ISSN
2331-8422
