Why Do Deep Residual Networks Generalize Better
than Deep Feedforward Networks? — A Neural
Tangent Kernel Perspective
Kaixuan Huang Yuqing Wang
Peking University Georgia Institute of Technology
hackyhuang@pku.edu.cn ywang3398@gatech.edu
Molei Tao Tuo Zhao
Georgia Institute of Technology Georgia Institute of Technology
mtao@gatech.edu tourzhao@gate ...
附件列表