Statistics
2 Questions
0 Answers
RANK
281,571
of 300,584
REPUTATION
0
CONTRIBUTIONS
2 Questions
0 Answers
ANSWER ACCEPTANCE
50.0%
VOTES RECEIVED
0
RANK
of 21,030
REPUTATION
N/A
AVERAGE RATING
0.00
CONTRIBUTIONS
0 Files
DOWNLOADS
0
ALL TIME DOWNLOADS
0
RANK
of 169,723
CONTRIBUTIONS
0 Problems
0 Solutions
SCORE
0
NUMBER OF BADGES
0
CONTRIBUTIONS
0 Posts
CONTRIBUTIONS
0 Public Channels
AVERAGE RATING
CONTRIBUTIONS
0 Highlights
AVERAGE NO. OF LIKES
Feeds
Question
运用MAPPO训练后进行验证，发现保存的所有Agent.mat文件的的验证reward回合奖励都是一个数，而且效果非常差
我运用MAPPO强化学习做了500个episode的训练，训练结果从奖励函数来看是收敛了，训练结果如下： 训练加载的环境相对于验证环境复杂一点，但是我验证时发现所有保存的Agent输出的回合奖励非常差，而且所有episode保存的Agent都是一样的...
2 years ago | 1 answer | 0
1
answerQuestion
PPO强化学习时动作空间超出设定的数据范围
在运用PPO方法进行强化学习训练时，我的动作空间设定如下： numAct = 1; actInfo = rlNumericSpec([numAct 1],"LowerLimit",-1,"UpperLimit",1); actInfo.Name = ...
2 years ago | 1 answer | 0
