Benign overfitting for maximum-margin linear classification holds under much weaker distributional assumptions than previously known, and the noisy strong-signal regime shows a new phase transition.
Thus, by Lemma S2.5 (i), ๐กโค ๐ผ2 2 1โ๐โฅ ๐โฅ2 <โฅ ๐โฅ2
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
Universality of Benign Overfitting in Binary Linear Classification
Benign overfitting for maximum-margin linear classification holds under much weaker distributional assumptions than previously known, and the noisy strong-signal regime shows a new phase transition.