4 pts) Sometimes MDPs are formulated with a reward functionR(s; a) that depends on theaction taken or with a reward functionR(s; a; s0) that also depends on the outcome state.(a) Write the Bellman equations for these formulations and show how an MDP with rewardfunctionR(s; a; s0) can be transformed into a di erent MDP with rewardR(s; a) such thatthe optimal policy in the new MDP corresponds exactly to the optimal policy in the originalMDP.(b) Do the same to convert an MDP withR(s; a) into an MDP withR(s).