A Combinatorial Algorithm for the Minimum (2, r)-Metric Problem and Some Generalizations
be a graph with nonnegative integer capacities c(e) of the edges \(\), and let μ be a metric that establishes distances on the pairs of elements of a subset \(\). In the minimum 0-extension problem (*), one is asked for finding a (semi)metric m on V such that m coincides with μ within T, each \(\) is at zero distance from some \(\), and the value \(\) is as small as possible. This is the classical minimum (undirected) cut problem when \(\) and \(\), and the minimum (2, r)-metric problem when μ is the path metric of the complete bipartite graph \(\). It is known that the latter problem can be solved in strongly polynomial time by use of the ellipsoid method.
We develop a polynomial time algorithm for the minimum (2, r)-metric problem, using only ``purely combinatorial'' means. The algorithm simultaneously solves a certain associated integer multiflow problem. We then apply this algorithm to solve (*) for a wider class of metrics μ, present other results and raise open questions.
Unable to display preview. Download preview PDF.