Two proofs for Shallow Packings - Archive ouverte HAL
Article Dans Une Revue Discrete and Computational Geometry Année : 2016

Two proofs for Shallow Packings

Résumé

We refine the bound on the packing number, originally shown by Haussler, for shallow geometric set systems. Specifically, let V be a finite set system defined over an n-point set X; we view V as a set of indicator vectors over the n-dimensional unit cube. A δ-separated set of V is a subcollection W, s.t. the Hamming distance between each pair u, v ∈ W is greater than δ, where δ > 0 is an integer parameter. The δ-packing number is then defined as the cardinality of a largest δ-separated subcollection of V. Haussler showed an asymptotically tight bound of Θ((n/δ) d) on the δ-packing number if V has VC-dimension (or primal shatter dimension) d. We refine this bound for the scenario where, for any subset, X ⊆ X of size m ≤ n and for any parameter 1 ≤ k ≤ m, the number of vectors of length at most k in the restriction of V to X is only O(m d1 k d−d1), for a fixed integer d > 0 and a real parameter 1 ≤ d 1 ≤ d (this generalizes the standard notion of bounded primal shatter dimension when d 1 = d). In this case when V is " k-shallow " (all vector lengths are at most k), we show that its δ-packing number is O(n d1 k d−d1 /δ d), matching Haussler's bound for the special cases where d 1 = d or k = n. We present two proofs, the first is an extension of Haussler's approach, and the second extends the proof of Chazelle, originally presented as a simplification for Haussler's proof.
Fichier principal
Vignette du fichier
DCG-SoCG15 (1).pdf (516.7 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01360460 , version 1 (16-09-2016)

Licence

Identifiants

Citer

Kunal Dutta, Esther Ezra, Arijit Ghosh. Two proofs for Shallow Packings . Discrete and Computational Geometry, 2016, Special Issue: 31st Annual Symposium on Computational Geometry, ⟨10.1007/s00454-016-9824-0⟩. ⟨hal-01360460⟩
536 Consultations
187 Téléchargements

Altmetric

Partager

More