Nonparametric estimations of non-negative random variables distributions

Vávra, František, et al. "Nonparametric estimations of non-negative random variables distributions." Kybernetika 39.3 (2003): [341]-346. <http://eudml.org/doc/33648>.

@article{Vávra2003,
abstract = {The problem of estimation of distribution functions or fractiles of non- negative random variables often occurs in the tasks of risk evaluation. There are many parametric models, however sometimes we need to know also some information about the shape and the type of the distribution. Unfortunately, classical approaches based on kernel approximations with a symmetric kernel do not give any guarantee of non-negativity for the low number of observations. In this note a heuristic approach, based on the assumption that non-negative distributions can be also approximated by means of kernels which are defined only on the positive real numbers, is discussed.},
author = {Vávra, František, Nový, Pavel, Mašková, Hana, Kotlíková, Michala, Zmrhal, David},
journal = {Kybernetika},
keywords = {distribution function; kernelapproximation; non-negative random variable; distribution function; kernel approximation},
language = {eng},
number = {3},
pages = {[341]-346},
publisher = {Institute of Information Theory and Automation AS CR},
title = {Nonparametric estimations of non-negative random variables distributions},
url = {http://eudml.org/doc/33648},
volume = {39},
year = {2003},
}

TY - JOUR
AU - Vávra, František
AU - Nový, Pavel
AU - Mašková, Hana
AU - Kotlíková, Michala
AU - Zmrhal, David
TI - Nonparametric estimations of non-negative random variables distributions
JO - Kybernetika
PY - 2003
PB - Institute of Information Theory and Automation AS CR
VL - 39
IS - 3
SP - [341]
EP - 346
AB - The problem of estimation of distribution functions or fractiles of non- negative random variables often occurs in the tasks of risk evaluation. There are many parametric models, however sometimes we need to know also some information about the shape and the type of the distribution. Unfortunately, classical approaches based on kernel approximations with a symmetric kernel do not give any guarantee of non-negativity for the low number of observations. In this note a heuristic approach, based on the assumption that non-negative distributions can be also approximated by means of kernels which are defined only on the positive real numbers, is discussed.
LA - eng
KW - distribution function; kernelapproximation; non-negative random variable; distribution function; kernel approximation
UR - http://eudml.org/doc/33648
ER -