1 paper · 1 filter
Paweł Piwek, Adam Klukowski, Tianyang Hu
Classic learning theory suggests that proper regularization is the key to good generalization and robustness. In classification, current training schemes only target the complexity…