I have a historical data (from real sensors) which have enough knowledge of the actions and states needed for the use of reinforcement learning and a modeled virtual environment of a real system (assuming it is good enough, but a reduced order model). How should I train a policy using DDPG using these in the shortest time possible with a relatively good result?

The resulting trained policy will then be tested on the real system and then be trained and used later online in the real system. It would be so good to have a relatively good policy before testing on the real system.


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.