Found 1 document(s)
Page 1 of 1
-
In reinforcement learning, one presumes the action set to be known ahead of time, and, thus, one must know if any state-action pairs are not permissible ahead of computations. In some applications, determining this actio...