TY  - RPRT
TI  - A Newton-Based Method for Nonconvex Optimization with Fast Evasion of Saddle Points
AU  - Santiago Paternain
AU  - Aryan Mokhtari
AU  - Alejandro Ribeiro
PY  - 2018
UR  - https://arxiv.org/abs/1707.08028
ID  - 1707.08028
ER  -