f*******g 发帖数: 55 | 1 Consider a deterministic infinite-horizon average cost problem.
Given a policy u, the state will converge to x0. Then in order to verify the
optimality of u, do we have to check the optimality equation (Bellman's
equation) for all the states in the state space, or do we just need to check
the equation for x0?
Thank you. |
|