DeepQLearning.jl
DeepQLearning.jl copied to clipboard
Support of AbtractEnvironment
This solver uses some function that are broader than the minimal interface defined in RLInterface and relies on internal fields such as env.problem
in many places.
Ideally, the solver should support an RL environment defined just using RLInterface.jl and without necessarily an MDP
or POMDP
object associated with it.
Yes, this is definitely important. In my class, more students had success with this package than any other, but this made it a little confusing to use.
Right now it is really designed to work with POMDPs.jl. Any AbtractEnvironment
could technically be implemented using MDP
and the generative interface.
initialstate
, gen
and actions
is all what's needed I believe.