Embedding Distances into the Hamming Cube
Résumé
In Coding Theory, two different decision criteria types are mostly used: a maximal likelihood (relative to a probabilistic model of a channel) and a nearest neighbour criterion (relative to a distance model of the channel). In this work we present an algorithm that, given a maximal likelihood criterion, decides if there is a nearest neighbour criterion that matches it and, in this case, produces a metric that determines such a criterion. It is also shown that the Hamming metric is universal, in the sense that any metric, up to a decoding equivalence, can be isometrically embedded into a hypercube with the Hamming metric.
Domaines
Théorie de l'information [cs.IT]Origine | Fichiers produits par l'(les) auteur(s) |
---|
Loading...