The notion of Vapnik-Chervonenkis (VC) class of sets can be extended to classes of functions in a few ways. Under further hypotheses, central limit theorems for empirical measures can be proved uniformly over such classes. Specific such classes on Euclidean spaces can be used to show the existence of location vector and scatter matrix functionals, replacing mean vectors and covariance matrices, but on classes of probability measures P that are weakly dense, weakly open, and so contain arbitrarily heavy-tailed distributions.
CITATION STYLE
Dudley, R. M. (2015). Classes of functions related to VC properties. In Measures of Complexity: Festschrift for Alexey Chervonenkis (pp. 185–196). Springer International Publishing. https://doi.org/10.1007/978-3-319-21852-6_14
Mendeley helps you to discover research relevant for your work.