What is the optimal value function for a one-step horizon for s21?

For a given MDP, there is a finite value of N such that the estimate of the optimal value function computed by the policy iteration algorithm after N iterations is the same as the estimate after N+1 iterations, to arbitrary precision.

Select all true statements: For a system with a linear motion model and a nonlinear but differentiable observation likelihood function, a state estimator can be implemented using:

What is the optimal value function for a one-step horizon for s24?

The total number of possible policies in an MDP with discrete states and actions is finite.

For the following questions, what is the optimal value function for a two-step horizon with a discount factor of 1? [Note that the optimal value function for a two-step horizon is optimal sum of the (discounted) rewards after taking two steps]

What is the optimal value function for a one-step horizon for s14?

Is the L1 norm an admissible heuristic on this graph?

Breadth-first search:

Consider the game search tree with true utilities indicated at each leaf: