TY  - RPRT
TI  - Knowledge Distillation in Wide Neural Networks: Risk Bound, Data Efficiency and Imperfect Teacher
AU  - Guangda Ji
AU  - Zhanxing Zhu
PY  - 2020
UR  - https://arxiv.org/abs/2010.10090
ID  - 2010.10090
ER  -