For a model of Markov Decision Process, Policy creation via two methods : Value Iteration and Linear Programming
Model world has 4*4 block grid, one positive terminal state, one negative terminal state.
is total description of the world.
contains the code which runs value iteration algorithm to find the utilities of all states and then final policy. It prints the result of every iteration.
has the same code but with different world model than given problem statement.
- It is another approach to find policy.
has the linear solver's output.
- Final output of the solver gives expected utility of start state and policy for the world.