Click here to flash read.
We study the asymptotic overfitting behavior of interpolation with minimum
norm ($\ell_2$ of the weights) two-layer ReLU networks for noisy univariate
regression. We show that overfitting is tempered for the $L_1$ loss, and any
$L_p$ loss for $p<2$, but catastrophic for $p\geq 2$.
Click here to read this post out
ID: 301662; Unique Viewers: 0
Voters: 0
Latest Change: July 31, 2023, 7:30 a.m.
Changes:
Dictionaries:
Words:
Spaces:
Comments:
Newcom