Title: Hierarchical-DQNr
Download

Description: There is such a bottleneck in the traditional table form of reinforcement learning. Table is used to store every state and Q value of every action in this state. However, today's problem is too complex. The state can be as many as the stars in the sky (compared with the following go). If all tables are used to store them, I'm afraid that our computer has not enough memory, and It is also a time-consuming task to search the corresponding status in such a large table
Downloaders recently:
[More information of uploader 小人物0104]]
To Search:
File list (Click to check if it's the file you need, and recomment it at the bottom):
文件名 | 大小 | 更新时间 |
---|---|---|
Hierarchical-DQN-master | 0 | 2017-11-23 |
Hierarchical-DQN-master\README.md | 19 | 2017-11-23 |
Hierarchical-DQN-master\dqn.py | 5905 | 2017-11-23 |
Hierarchical-DQN-master\hierarchical_dqn.py | 8131 | 2017-11-23 |
Hierarchical-DQN-master\replay_buffer.py | 2458 | 2017-11-23 |
Hierarchical-DQN-master\train_dqn.py | 6628 | 2017-11-23 |