Let x[SIZE="1"]1, . . . , x[SIZE="1"]k be feasible solutions of the linear programming problem:
Maximize z = (c^t)*x subject to Ax < = b and x >= 0,
so for i= 1, . . . , k, Ax[SIZE="1"]i <=b and x[SIZE="1"]i >=0,
Let v be any convex linear combination of x[SIZE="1"]1, . . . ...