TY  - RPRT
TI  - Gradient Descent Provably Optimizes Over-parameterized Neural Networks
AU  - Simon S. Du
AU  - Xiyu Zhai
AU  - Barnabas Poczos
AU  - Aarti Singh
PY  - 2019
UR  - https://arxiv.org/abs/1810.02054
ID  - 1810.02054
ER  -