paper

A class of Rényi information estimators for multidimensional densities

arXiv:0810.5302 · doi:10.1214/07-AOS539

Abstract

A class of estimators of the Rényi and Tsallis entropies of an unknown distribution in is presented. These estimators are based on the th nearest-neighbor distances computed from a sample of i.i.d. vectors with distribution . We show that entropies of any order , including Shannon's entropy, can be estimated consistently with minimal assumptions on . Moreover, we show that it is straightforward to extend the nearest-neighbor method to estimate the statistical distance between two distributions using one i.i.d. sample from each. (Wit Correction.)

Published in at http://dx.doi.org/10.1214/07-AOS539 the Annals of Statistics (http://www.imstat.org/aos/) by the Institute of Mathematical Statistics (http://www.imstat.org)

Cited by in corpus (47)