Customized Action Selection in RL DQN
3 Ansichten (letzte 30 Tage)
Ältere Kommentare anzeigen
Hi,
I would like to ask if the latest Reinforcement Learning (RL) toolbox version supports customized action selection.
I’m currently using a DQN agent, and the action in each time step is selected randomly following the epsilon-greedy algorithm. However, I would like to feed in some probabilities in the action selection, such that certain actions are more likely to be chosen. Is this possible using the RL toolbox?
Thank you!
0 Kommentare
Antworten (1)
Emmanouil Tzorakoleftherakis
am 16 Jan. 2021
Bearbeitet: Emmanouil Tzorakoleftherakis
am 16 Jan. 2021
Hello,
I believe this is not possible yet. A potential workaround (although not state dependent) would be to emulate a pdf by providing actions with higher probabilities multiple times when creating your action space with rlFinitesetSpec but I haven't tested that. So something like:
actInfo = rlFiniteSetSpec([-2 0 2 2 2])
1 Kommentar
Siehe auch
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!