a dynamic programming algorithm memorizes the solutions of optimal subproblems in an organized, tabular form (a dynamic programming matrix), so that each subproblem is solved just once.