Reinforcement Learning（四）：Actor-Critic Methods
分类:
文章
•
2025-02-24 23:16:11
主要思想：
Policy Network (Actor)
Value Network (Critic):
形象对比：
Train the Neural Networks
具体步骤：
Update value network q using TD
Update policy network Π using policy gradient
Actor-Critic Method
Summary of Algorithm
Summary
Policy Network and Value Network
Training
