Whenever you're ready, walk me through how you'd approach this.
Walk through the mathematical steps of a dynamic programming approach to solve a sequential decision-making problem.
Define a finite Markov decision process with states, actions, transition probabilities, rewards, and a discount factor. Derive the Bellman optimality equation, explain value iteration or policy iteration, and implement one method for a supplied transition model. Show how you would verify convergence and evaluate the resulting policy.