0
$\begingroup$

In most of the reinforcement learning literature, I see that there is a shaded area in the graphs. I couldn't understand what it exactly represents?

For example, from the A3C paper:

enter image description here

Or another example from the PPO paper:

enter image description here

Is it for multiple runs or it's for something else? How I can reproduce such graphs (which library and what type of data from my training episode do I need)?

$\endgroup$
1
$\begingroup$

They train the agent multiple times, then plot the mean +- standard deviation of the agent performance (the shaded region is representing the standard deviation).

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.