There are various meta-learning algorithms in RL that are proposed for settings when we have a (deep) neural network and the policy (or the value function) are parameterized as such.
Can these methods be applied to tabular cases? For instance by following the steps of algorithm 3 in the MAML paper (https://arxiv.org/pdf/1703.03400.pdf), or there is a theoretical obstacle that I am not aware of.
