Matches in SemOpenAlex for { <https://semopenalex.org/work/W2973842521> ?p ?o ?g. }
- W2973842521 abstract "Robust statistics traditionally focuses on outliers, or perturbations in total variation distance. However, a dataset could be corrupted in many other ways, such as systematic measurement errors and missing covariates. We generalize the robust statistics approach to consider perturbations under any Wasserstein distance, and show that robust estimation is possible whenever a distribution's population statistics are robust under a certain family of friendly perturbations. This generalizes a property called resilience previously employed in the special case of mean estimation with outliers. We justify the generalized resilience property by showing that it holds under moment or hypercontractive conditions. Even in the total variation case, these subsume conditions in the literature for mean estimation, regression, and covariance estimation; the resulting analysis simplifies and sometimes improves these known results in both population limit and finite-sample rate. Our robust estimators are based on minimum distance (MD) functionals (Donoho and Liu, 1988), which project onto a set of distributions under a discrepancy related to the perturbation. We present two approaches for designing MD estimators with good finite-sample rates: weakening the discrepancy and expanding the set of distributions. We also present connections to Gao et al. (2019)'s recent analysis of generative adversarial networks for robust estimation." @default.
- W2973842521 created "2019-09-26" @default.
- W2973842521 creator A5031613608 @default.
- W2973842521 creator A5034192173 @default.
- W2973842521 creator A5060196069 @default.
- W2973842521 date "2019-09-18" @default.
- W2973842521 modified "2023-09-27" @default.
- W2973842521 title "Generalized Resilience and Robust Statistics" @default.
- W2973842521 cites W1483679902 @default.
- W2973842521 cites W1509803206 @default.
- W2973842521 cites W1546372806 @default.
- W2973842521 cites W156419107 @default.
- W2973842521 cites W1585160083 @default.
- W2973842521 cites W1585566614 @default.
- W2973842521 cites W1968355947 @default.
- W2973842521 cites W1968956560 @default.
- W2973842521 cites W1985124715 @default.
- W2973842521 cites W2006228748 @default.
- W2973842521 cites W2008229822 @default.
- W2973842521 cites W2018000058 @default.
- W2973842521 cites W2025533349 @default.
- W2973842521 cites W2025720061 @default.
- W2973842521 cites W2029538739 @default.
- W2973842521 cites W2037468546 @default.
- W2973842521 cites W2039587747 @default.
- W2973842521 cites W2039892753 @default.
- W2973842521 cites W2042132284 @default.
- W2973842521 cites W2057712948 @default.
- W2973842521 cites W2063178190 @default.
- W2973842521 cites W2063543800 @default.
- W2973842521 cites W2065742895 @default.
- W2973842521 cites W2084238990 @default.
- W2973842521 cites W2102714321 @default.
- W2973842521 cites W2137130182 @default.
- W2973842521 cites W2141042406 @default.
- W2973842521 cites W2166481425 @default.
- W2973842521 cites W2338990760 @default.
- W2973842521 cites W2564590721 @default.
- W2973842521 cites W2589006171 @default.
- W2973842521 cites W2592318711 @default.
- W2973842521 cites W2592371997 @default.
- W2973842521 cites W2597655115 @default.
- W2973842521 cites W2726793048 @default.
- W2973842521 cites W2771719469 @default.
- W2973842521 cites W2773686537 @default.
- W2973842521 cites W2774423163 @default.
- W2973842521 cites W2787248994 @default.
- W2973842521 cites W2791820624 @default.
- W2973842521 cites W2806187986 @default.
- W2973842521 cites W2806196585 @default.
- W2973842521 cites W2895204589 @default.
- W2973842521 cites W2896534181 @default.
- W2973842521 cites W2920287530 @default.
- W2973842521 cites W2922488429 @default.
- W2973842521 cites W2942689850 @default.
- W2973842521 cites W2949006600 @default.
- W2973842521 cites W2949506549 @default.
- W2973842521 cites W2951315160 @default.
- W2973842521 cites W2955499441 @default.
- W2973842521 cites W2962826017 @default.
- W2973842521 cites W2963616494 @default.
- W2973842521 cites W2964135521 @default.
- W2973842521 cites W2964317125 @default.
- W2973842521 cites W2965497096 @default.
- W2973842521 cites W2970007412 @default.
- W2973842521 cites W2984272935 @default.
- W2973842521 cites W3013236627 @default.
- W2973842521 cites W3017335724 @default.
- W2973842521 cites W3021023808 @default.
- W2973842521 cites W3029013756 @default.
- W2973842521 cites W3039117700 @default.
- W2973842521 cites W3043502075 @default.
- W2973842521 cites W3045703125 @default.
- W2973842521 cites W628915549 @default.
- W2973842521 doi "https://doi.org/10.48550/arxiv.1909.08755" @default.
- W2973842521 hasPublicationYear "2019" @default.
- W2973842521 type Work @default.
- W2973842521 sameAs 2973842521 @default.
- W2973842521 citedByCount "18" @default.
- W2973842521 countsByYear W29738425212019 @default.
- W2973842521 countsByYear W29738425212020 @default.
- W2973842521 countsByYear W29738425212021 @default.
- W2973842521 crossrefType "posted-content" @default.
- W2973842521 hasAuthorship W2973842521A5031613608 @default.
- W2973842521 hasAuthorship W2973842521A5034192173 @default.
- W2973842521 hasAuthorship W2973842521A5060196069 @default.
- W2973842521 hasBestOaLocation W29738425211 @default.
- W2973842521 hasConcept C104317684 @default.
- W2973842521 hasConcept C105795698 @default.
- W2973842521 hasConcept C119043178 @default.
- W2973842521 hasConcept C144024400 @default.
- W2973842521 hasConcept C149923435 @default.
- W2973842521 hasConcept C178650346 @default.
- W2973842521 hasConcept C185429906 @default.
- W2973842521 hasConcept C185592680 @default.
- W2973842521 hasConcept C2908647359 @default.
- W2973842521 hasConcept C33923547 @default.
- W2973842521 hasConcept C55493867 @default.
- W2973842521 hasConcept C63479239 @default.
- W2973842521 hasConcept C67226441 @default.