In the context of the reinforcement learning domain, the dichotomy between model-based (learn a model and used it to determine a controller) and model-free (learn a controller without learning a model) approaches is a contested one.

There are many algorithms that have been developed for both approaches.

My question is, are there any approaches that are between model-based and model-free? One example could be to specify a controller that is the combination of a learned controller (when there is sufficient information to do so) and a pre-determined controller based a model (when we cannot learn the controller).



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.