Letter to the Editor-Solution of a Combinatorial Problem by Dynamic Programming

A combinatorial problem to determine the least cost of a matrix configuration is solved by dynamic programming. In particular, the application of an algorithm based on an approximation in policy space yields a least-cost configuration consistent with the initial matrix configuration. While the optimization is local in character because of its dependence on the initial matrix configuration, the optimization may be extended to approach global optimization by consideration of a variety of initial matrix configurations. The method lends itself readily to finding the hierachy of configurations ranging from the least to the most expensive.