Indexability, concentration, and VC theory
Data Structures and Algorithms
2012-04-13 v5 Machine Learning
Abstract
Degrading performance of indexing schemes for exact similarity search in high dimensions has long since been linked to histograms of distributions of distances and other 1-Lipschitz functions getting concentrated. We discuss this observation in the framework of the phenomenon of concentration of measure on the structures of high dimension and the Vapnik-Chervonenkis theory of statistical learning.
Keywords
Cite
@article{arxiv.1008.5105,
title = {Indexability, concentration, and VC theory},
author = {Vladimir Pestov},
journal= {arXiv preprint arXiv:1008.5105},
year = {2012}
}
Comments
17 pages, final submission to J. Discrete Algorithms (an expanded, improved and corrected version of the SISAP'2010 invited paper, this e-print, v3)