TY  - RPRT
TI  - Uniform Learning in a Deep Neural Network via "Oddball" Stochastic Gradient Descent
AU  - Andrew J. R. Simpson
PY  - 2015
UR  - https://arxiv.org/abs/1510.02442
ID  - 1510.02442
ER  -