I've been thinking about the question and I don't yet see how to connect gcForest with Reinforcement Learning and if it is actually possible. To be honest it is probably due to my limited knowledge about RL and will need to think longer about this.
I'll come back to you if I find a way.
in my opinion, gcForest is a supervised framework , it learns to fit a curve.
in deep RL such as deep q-learning, replace the MLP part may be a starting point to play with.
I've been thinking about the question and I don't yet see how to connect gcForest with Reinforcement Learning and if it is actually possible. To be honest it is probably due to my limited knowledge about RL and will need to think longer about this. I'll come back to you if I find a way.
By the way, what do you think ?