Matches in SemOpenAlex for { <https://semopenalex.org/work/W3020361330> ?p ?o ?g. }
- W3020361330 abstract "We prove bounds on the population risk of the maximum margin algorithm for two-class linear classification. For linearly separable training data, the maximum margin algorithm has been shown in previous work to be equivalent to a limit of training with logistic loss using gradient descent, as the training error is driven to zero. We analyze this algorithm applied to random data including misclassification noise. Our assumptions on the clean data include the case in which the class-conditional distributions are standard normal distributions. The misclassification noise may be chosen by an adversary, subject to a limit on the fraction of corrupted labels. Our bounds show that, with sufficient over-parameterization, the maximum margin algorithm trained on noisy data can achieve nearly optimal population risk." @default.
- W3020361330 created "2020-05-01" @default.
- W3020361330 creator A5033738047 @default.
- W3020361330 creator A5084487318 @default.
- W3020361330 date "2020-04-25" @default.
- W3020361330 modified "2023-09-27" @default.
- W3020361330 title "Finite-sample Analysis of Interpolating Linear Classifiers in the Overparameterized Regime" @default.
- W3020361330 cites W1530699444 @default.
- W3020361330 cites W1535793628 @default.
- W3020361330 cites W1542886316 @default.
- W3020361330 cites W1554944419 @default.
- W3020361330 cites W1930161385 @default.
- W3020361330 cites W1964923438 @default.
- W3020361330 cites W1977696979 @default.
- W3020361330 cites W1982381767 @default.
- W3020361330 cites W1995897489 @default.
- W3020361330 cites W1997988698 @default.
- W3020361330 cites W2029394273 @default.
- W3020361330 cites W2049701820 @default.
- W3020361330 cites W2051605894 @default.
- W3020361330 cites W2063698478 @default.
- W3020361330 cites W2063844956 @default.
- W3020361330 cites W2064921494 @default.
- W3020361330 cites W2098933520 @default.
- W3020361330 cites W2106458073 @default.
- W3020361330 cites W2106491486 @default.
- W3020361330 cites W2107189314 @default.
- W3020361330 cites W2112876447 @default.
- W3020361330 cites W2113290770 @default.
- W3020361330 cites W2118250684 @default.
- W3020361330 cites W2127265926 @default.
- W3020361330 cites W2161278885 @default.
- W3020361330 cites W2163614729 @default.
- W3020361330 cites W2186768631 @default.
- W3020361330 cites W2490901831 @default.
- W3020361330 cites W2536321907 @default.
- W3020361330 cites W2616625443 @default.
- W3020361330 cites W2620857485 @default.
- W3020361330 cites W2621358189 @default.
- W3020361330 cites W2752860283 @default.
- W3020361330 cites W2752885492 @default.
- W3020361330 cites W2787248994 @default.
- W3020361330 cites W2799061466 @default.
- W3020361330 cites W2803953503 @default.
- W3020361330 cites W2806252860 @default.
- W3020361330 cites W2899476926 @default.
- W3020361330 cites W2911742574 @default.
- W3020361330 cites W2923764619 @default.
- W3020361330 cites W2962703558 @default.
- W3020361330 cites W2963038274 @default.
- W3020361330 cites W2963060833 @default.
- W3020361330 cites W2963094815 @default.
- W3020361330 cites W2963518130 @default.
- W3020361330 cites W2963742538 @default.
- W3020361330 cites W2963798163 @default.
- W3020361330 cites W2963826371 @default.
- W3020361330 cites W2964063947 @default.
- W3020361330 cites W2965772785 @default.
- W3020361330 cites W2967536008 @default.
- W3020361330 cites W2970290137 @default.
- W3020361330 cites W2970584880 @default.
- W3020361330 cites W2989169642 @default.
- W3020361330 cites W3004510491 @default.
- W3020361330 cites W3007237875 @default.
- W3020361330 cites W3014316192 @default.
- W3020361330 cites W3015944005 @default.
- W3020361330 cites W3018252856 @default.
- W3020361330 cites W3021023808 @default.
- W3020361330 cites W3025991882 @default.
- W3020361330 cites W3046680711 @default.
- W3020361330 cites W3046705928 @default.
- W3020361330 cites W3084611071 @default.
- W3020361330 cites W3099518927 @default.
- W3020361330 cites W3101595931 @default.
- W3020361330 cites W3104969455 @default.
- W3020361330 cites W3111350549 @default.
- W3020361330 cites W3121962434 @default.
- W3020361330 cites W3132591740 @default.
- W3020361330 cites W3137695714 @default.
- W3020361330 cites W3157731489 @default.
- W3020361330 cites W3168158992 @default.
- W3020361330 cites W369786348 @default.
- W3020361330 cites W740415 @default.
- W3020361330 hasPublicationYear "2020" @default.
- W3020361330 type Work @default.
- W3020361330 sameAs 3020361330 @default.
- W3020361330 citedByCount "0" @default.
- W3020361330 crossrefType "posted-content" @default.
- W3020361330 hasAuthorship W3020361330A5033738047 @default.
- W3020361330 hasAuthorship W3020361330A5084487318 @default.
- W3020361330 hasConcept C105795698 @default.
- W3020361330 hasConcept C11413529 @default.
- W3020361330 hasConcept C115961682 @default.
- W3020361330 hasConcept C119857082 @default.
- W3020361330 hasConcept C134306372 @default.
- W3020361330 hasConcept C144024400 @default.
- W3020361330 hasConcept C149923435 @default.
- W3020361330 hasConcept C151201525 @default.
- W3020361330 hasConcept C153180895 @default.
- W3020361330 hasConcept C154945302 @default.