I mean how we can know a whether a "model-based reinforcement learning algorithm" or a "model-free reinforcement learning algorithm" is suitable for our case . Furthermore, there are a lot of algorithms to choose in each category (i.e., model-based or model-free), how we can find the most suitable algorithm. For example how we can choose between Q-learning, SARSA or TD-learning?
It is always a challenge to choose an appropriate algorithm to solve any problems. You can try all of them and then decide.
From you question, I am not sure what is "your case". However what ever problem you are solving, you could let the computer choosing the best algorithms for different instances. Then you will be able to decide which ones suit best your needs.
To do so, I would use some hyper-heuristics techniques or some programming optimisation used by Holger Hoos.
I hope it helps and good luck in your study.
Patricia
Interesting Amir.
I think both of our answer complement each others.
Patricia
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence