For optimization linear programming class, could you explain what you mean by drawing a subproblem tree and how it relates to the optimization process?