I am trying to write a paper for my optimization class about Approximate Dynamic Programming. I found a few good papers but they all seem to dive straight into the material without talking about the basics so I am really lost about what it is.

In class we talked a little bit about "Dynamic Programming" but I can't seem to find out the difference between regular DP and ADP. What are we approximating? One source says

Many problems in these fields are described by continuous variables, whereas DP and RL can find exact solutions only in the discrete case. Therefore, approximation is essential in practical DP and RL.

so I thought ADP is used for continuous time but then another paper states

Approximate Dynamic Programming (ADP) is a powerful technique to solve large scale discrete time multistage stochastic control processes

So is it for discrete time or continuous? Or both?

Similarly, a lot of sources on ADP talk about reinforcement learning but they don't distinguish the two. Are they both both pretty much the same things or is there a distinct difference?

  • $\begingroup$ It might be useful to cite the names of the papers from which you have taken these excerpts. $\endgroup$ – nbro May 10 at 22:33

Your Answer

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

Browse other questions tagged or ask your own question.