FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385463474> ?p ?o ?g. }

• W4385463474 abstract "This paper considers the problem of supervised learning with linear methods when both features and labels can be corrupted, either in the form of heavy tailed data and/or corrupted rows. We introduce a combination of coordinate gradient descent as a learning algorithm together with robust estimators of the partial derivatives. This leads to robust statistical learning methods that have a numerical complexity nearly identical to non-robust ones based on empirical risk minimization. The main idea is simple: while robust learning with gradient descent requires the computational cost of robustly estimating the whole gradient to update all parameters, a parameter can be updated immediately using a robust estimator of a single partial derivative in coordinate gradient descent. We prove upper bounds on the generalization error of the algorithms derived from this idea, that control both the optimization and statistical errors with and without a strong convexity assumption of the risk. Finally, we propose an efficient implementation of this approach in a new Python library called linlearn, and demonstrate through extensive numerical experiments that our approach introduces a new interesting compromise between robustness, statistical performance and numerical efficiency for this problem." @default.
• W4385463474 created "2023-08-02" @default.
• W4385463474 creator A5060452597 @default.
• W4385463474 creator A5081026572 @default.
• W4385463474 date "2023-08-01" @default.
• W4385463474 modified "2023-10-02" @default.
• W4385463474 title "Robust supervised learning with coordinate gradient descent" @default.
• W4385463474 cites W1514928307 @default.
• W4385463474 cites W1753414968 @default.
• W4385463474 cites W1969811075 @default.
• W4385463474 cites W1971334823 @default.
• W4385463474 cites W1978239142 @default.
• W4385463474 cites W1984332158 @default.
• W4385463474 cites W2000769684 @default.
• W4385463474 cites W2002344122 @default.
• W4385463474 cites W2006398000 @default.
• W4385463474 cites W2022851810 @default.
• W4385463474 cites W2023901033 @default.
• W4385463474 cites W2040324983 @default.
• W4385463474 cites W2046033161 @default.
• W4385463474 cites W2049058890 @default.
• W4385463474 cites W2072925128 @default.
• W4385463474 cites W2073388398 @default.
• W4385463474 cites W2076541756 @default.
• W4385463474 cites W2080745194 @default.
• W4385463474 cites W2085261163 @default.
• W4385463474 cites W2091825929 @default.
• W4385463474 cites W2095225664 @default.
• W4385463474 cites W2095984592 @default.
• W4385463474 cites W2097452095 @default.
• W4385463474 cites W2124541940 @default.
• W4385463474 cites W2138019504 @default.
• W4385463474 cites W2142971854 @default.
• W4385463474 cites W2145962650 @default.
• W4385463474 cites W2152734820 @default.
• W4385463474 cites W2161134697 @default.
• W4385463474 cites W2166183437 @default.
• W4385463474 cites W2182886880 @default.
• W4385463474 cites W2288274632 @default.
• W4385463474 cites W2489822048 @default.
• W4385463474 cites W2554864439 @default.
• W4385463474 cites W2585502432 @default.
• W4385463474 cites W2592371997 @default.
• W4385463474 cites W2621365674 @default.
• W4385463474 cites W2806473647 @default.
• W4385463474 cites W2942689850 @default.
• W4385463474 cites W2949006600 @default.
• W4385463474 cites W2962730199 @default.
• W4385463474 cites W2962927523 @default.
• W4385463474 cites W2963308017 @default.
• W4385463474 cites W2963351358 @default.
• W4385463474 cites W2963368181 @default.
• W4385463474 cites W2963616494 @default.
• W4385463474 cites W2965685971 @default.
• W4385463474 cites W3010140118 @default.
• W4385463474 cites W3017335724 @default.
• W4385463474 cites W3019929158 @default.
• W4385463474 cites W3032338640 @default.
• W4385463474 cites W3032696844 @default.
• W4385463474 cites W3035618875 @default.
• W4385463474 cites W3100743579 @default.
• W4385463474 cites W3104898220 @default.
• W4385463474 cites W3111890340 @default.
• W4385463474 cites W3126729338 @default.
• W4385463474 cites W4206742934 @default.
• W4385463474 cites W4211015534 @default.
• W4385463474 cites W4243072198 @default.
• W4385463474 cites W4245577611 @default.
• W4385463474 cites W4301425558 @default.
• W4385463474 doi "https://doi.org/10.1007/s11222-023-10283-7" @default.
• W4385463474 hasPublicationYear "2023" @default.
• W4385463474 type Work @default.
• W4385463474 citedByCount "0" @default.
• W4385463474 crossrefType "journal-article" @default.
• W4385463474 hasAuthorship W4385463474A5060452597 @default.
• W4385463474 hasAuthorship W4385463474A5081026572 @default.
• W4385463474 hasBestOaLocation W43854634742 @default.
• W4385463474 hasConcept C104317684 @default.
• W4385463474 hasConcept C105795698 @default.
• W4385463474 hasConcept C106159729 @default.
• W4385463474 hasConcept C11413529 @default.
• W4385463474 hasConcept C126255220 @default.
• W4385463474 hasConcept C153258448 @default.
• W4385463474 hasConcept C154945302 @default.
• W4385463474 hasConcept C157553263 @default.
• W4385463474 hasConcept C162324750 @default.
• W4385463474 hasConcept C185429906 @default.
• W4385463474 hasConcept C185592680 @default.
• W4385463474 hasConcept C33923547 @default.
• W4385463474 hasConcept C41008148 @default.
• W4385463474 hasConcept C50644808 @default.
• W4385463474 hasConcept C55493867 @default.
• W4385463474 hasConcept C63479239 @default.
• W4385463474 hasConcept C72134830 @default.
• W4385463474 hasConceptScore W4385463474C104317684 @default.
• W4385463474 hasConceptScore W4385463474C105795698 @default.
• W4385463474 hasConceptScore W4385463474C106159729 @default.
• W4385463474 hasConceptScore W4385463474C11413529 @default.
• W4385463474 hasConceptScore W4385463474C126255220 @default.
• W4385463474 hasConceptScore W4385463474C153258448 @default.

• next