Asymmetric distances, semidirected networks and majority in Fermat–Weber problems

Abstract

The Fermat-Weber problem is considered with distance defined by a quasimetric, an asymmetric and possibly nondefinite generalisation of a metric. In such a situation a distinction has to be made between sources and destinations. We show how the classical result of optimality at a destination or a source with majority weight, valid in a metric space, may be generalized to certain quasimetric spaces. The notion of majority has however to be strengthened to supermajority, defined by way of a measure of the asymmetry of the distance, which should be finite. This extended majority theorem applies to most asymmetric distance measures previously studied in literature, since these have finite asymmetry measure.Perhaps the most important application of quasimetrics arises in semidirected networks, which may contain edges of different (possibly zero) length according to direction, or directed edges. Distance in a semidirected network does not necessarily have finite asymmetry measure. But it is shown that an adapted majority result holds nevertheless in this important context, thanks to an extension of the classical node-optimality result to semidirected networks with constraints.Finally the majority theorem is further extended to Fermat-Weber problems with mixed asymmetric distances.

Locating a median line with partial coverage distance

“…Theorem 6 settles the question for norm distances d. In order to prove it, we need a generalized version of the Cauchy-Schwarz inequality which can be found for example in [14,22,23].…”

“…Observe that the first constraint is equivalent to two linear constraints, since we have by the generalized Cauchy-Schwarz inequality (see again [14,22,23])…”

On turbulent relations

“…For a relation given by the action of a group G on a space X, the local equivalence classes are called local orbits in Hjorth [5], and the notation (14). Similarly, for a uniform relation induced by a generalized pseudo-metric d on a set X, the notation (15).…”

A variational inequality approach for constrained multifacility Weber problem under gauge

“…In spite of the pioneering work of [33], the asymmetric distance has started to attract the researchers' interest until the recent several decades, and some progress has been made in both its theoretical and computational aspects recently, see e.g. [2,3,7,15,22,24].…”

“…For more details about the gauge and dual gauge, as well as their properties, the readers can be referred to [24]. According to (10), CMFWP (7) is equivalent to the following min-max problem:…”