There are some problems: mismatch of model and reality; gradient explosionThere are some p
阅读背景：
CS294-112深度增强学习课程（加州大学伯克利分校 2017）NO.4 Learning policies by imitating optimal controllers
来源：互联网
分享到：
