策略梯度方法

1. Basic idea of policy gradient2. Metrics to define optimal policies3. Gradients of the metrics4. Gradient-ascent algorithm (REINFORCE)参考文献本文是一篇学习笔记内容全部源自于以下视频https://www.bilibili.com/video/BV1Pz5C6iE3X/?p6spm_id_from333.1007.top_right_bar_window_history.content.clickvd_source44ed90827c8f67247cab0ab288133c80