Abstract:
A method is proposed for design of a sequence of approximate solutions of the linear programming problems in the functional space. This sequence is proved to converge to an optimal solution. The method makes use of the relation between the problem and a game in the functional space.