New answers tagged

1 vote

Why random sample from replay for DQN?

My intuition would tell me the sequence is what is most important in reinforcement learning. Most episodes have a delayed reward so most action/states do not have a reward (and are not "...
  • 733
0 votes

RL Advantage function why A = Q-V instead of A=V-Q?

๐‘„(๐‘ ,๐‘Ž) Q ( s , a ) returns a value of entire total reward that's expected ultimately, after we pick an action ๐‘Ž a . ๐‘‰(๐‘ ) V ( s ) is the same, just with an extra reward from current state ๐‘  s ...
  • 733

Top 50 recent answers are included