Based on former papers about valuated (Delta-)matroids we introduce the concept of a valuation of a graph. It is proved that in valuated Hamming graphs, a best solution as well as a 2-best solution with a given distance to the best solution can be reached via local research on shortest paths. Valuations of cubes are classified.
Keywords:
Combinatorial optimization, distance constraints, valuated (Delta-)matroids, Hamming graphs, cubes, convexity in graphs, local search, concave functions