Formal Description of Dynamic Programming (DP)

Formal Description​

si - denotes the state of the system at ith time, for all i = 1, 2, …, n (state transitions).

xi - denotes the decision made at ith time for all i = 1, 2, …, n.

                   formal representation

                                                           Figure 1

In other words, for all i = 1,2,…,n-1, the state transitions are denoted as:

                   formal representation

Let the cost of the changing state from si to si=1 be denoted as gi(si,xi)

Then the total cost is given as:

                   formal representation

Design Methodology for DP

Four step method:

  1. Characterize the Bellman equation for the problem.
  2. Recursively find the optimal solution and itscorresponding value for each tail problem.
  3. Compute the value of an optimal solution in a bottom-upfashion.
  4. Construct an optimal solution from computedinformation.

              Nikola Andrić