https://zhuanlan.zhihu.com/p/22810533
L2 -> Regression Problems
KL -> Classification Problems
http://deeplearning.cs.cmu.edu/slides/lec8.stochastic_gradient.pdf
CMU Deep Learning 2018 by Bhiksha Raj 学习记录(7)
原文:https://www.cnblogs.com/ecoflex/p/8886201.html