Kantorovich-Rubinstein Quasi-Metrics II: Hyperspaces and Powerdomains - Inria - Institut national de recherche en sciences et technologies du numérique
Article Dans Une Revue Topology and its Applications Année : 2022

Kantorovich-Rubinstein Quasi-Metrics II: Hyperspaces and Powerdomains

Résumé

We show that the Kantorovich-Rubinstein quasi-metrics d_KR and d^a_KR of Part I extend naturally to various spaces of previsions, and in particular not just the linear previsions (roughly, measures) of Part I. There are natural isomorphisms between the Hoare and Smyth powerdomains, as used in denotational semantics, and spaces of discrete sublinear previsions, and of discrete normalized superlinear previsions, respectively. Turning to the corresponding hyperspaces, namely the same powerdomains, but equipped with the lower Vietoris and upper Vietoris topologies instead, this turns into homeomorphisms with the corresponding space of previsions, equipped with the so-called weak topology. Through these isomorphisms again, the two powerdomains inherit quasi-metrics d_H and d_Q , respectively, that are reminiscent of the well-known Hausdorff metric. Then we show that the Hoare and Smyth powerdomains of an algebraic complete quasi-metric space are again algebraic complete, with those quasi-metrics, and similarly that the corresponding powerdomains of continuous complete quasi-metric spaces are continuous complete. Furthermore, in the continuous complete case, the d_H-Scott topology coincides with the lower Vietoris topology, and the d_Q-Scott topology coincides with the upper Vietoris topology.
Fichier principal
Vignette du fichier
distpdom5.pdf (528.79 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03383036 , version 1 (18-10-2021)

Identifiants

Citer

Jean Goubault-Larrecq. Kantorovich-Rubinstein Quasi-Metrics II: Hyperspaces and Powerdomains. Topology and its Applications, 2022, 305, pp.107885. ⟨10.1016/j.topol.2021.107885⟩. ⟨hal-03383036⟩
100 Consultations
87 Téléchargements

Altmetric

Partager

More