Generalizability of Neural Networks Minimizing Empirical Risk Based on Expressive Power.

Lijia Yu, Yibo Miao, Yifan Zhu, Xiao-Shan Gao, Lijun Zhang

Browse the full ICLR paper archive.