Robustly Learning a Gaussian: Getting Optimal Error, Efficiently

We study the fundamental problem of learning the parameters of a high-dimensional Gaussian in the presence of noise | where an "-fraction of our samples were chosen by an adversary. We give robust estimators that achieve estimation error O(ϵ) in the total variation distance, which is optimal u...

ver descrição completa

Detalhes bibliográficos
Principais autores: Stewart, Alistair, Diakonikolas, Ilias, Kamath, Gautam Chetan, Kane, Daniel M, Li, Jerry Zheng, Moitra, Ankur
Outros Autores: Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory
Formato: Artigo
Publicado em: Society for Industrial and Applied Mathematics 2018
Acesso em linha:http://hdl.handle.net/1721.1/116214
https://orcid.org/0000-0003-0048-2559
https://orcid.org/0000-0002-9937-0049
https://orcid.org/0000-0001-7047-0495