Detail publikace

Statistical Methods for Analyzing Musk Compounds Concentration based on Doubly Left-Censored Samples

FUSEK, M. MICHÁLEK, J.

Originální název

Statistical Methods for Analyzing Musk Compounds Concentration based on Doubly Left-Censored Samples

Typ

článek v časopise - ostatní, Jost

Jazyk

angličtina

Originální abstrakt

This contribution is focused on statistical methods for analyzing the worldwide commonly used synthetic musk compounds. Method of maximum likelihood considering doubly left-censored samples is used for statistical modeling of musk compound concentration. As for model distributions, the exponential and Weibull distributions are considered. The suitability of replacement of Weibull distribution with exponential distribution is explored using the asymptotic tests (Lagrange multiplier test, likelihood ratio test, Wald test). Moreover, using the asymptotic properties of maximum likelihood estimates, methods for comparison of two censored samples from exponential distribution are proposed and applied in analysis of concentrations of musk compounds extracted from the fish samples caught in front of and behind a wastewater treatment plant. The power functions of particular tests are compared by simulations.

Klíčová slova

Musk compounds, maximum likelihood, doubly left-censored sample, Weibull distribution, exponential distribution, Lagrange multiplier, likelihood ratio, Wald test

Autoři

FUSEK, M.; MICHÁLEK, J.

Rok RIV

2013

Vydáno

25. 10. 2013

ISSN

1998-0140

Periodikum

INTERNATIONAL JOURNAL of MATHEMATICAL MODELS AND METHODS IN APPLIED SCIENCES

Ročník

7

Číslo

8

Stát

Spojené státy americké

Strany od

755

Strany do

763

Strany počet

9

BibTex

@article{BUT102153,
  author="Michal {Fusek} and Jaroslav {Michálek}",
  title="Statistical Methods for Analyzing Musk Compounds Concentration based on Doubly Left-Censored Samples",
  journal="INTERNATIONAL JOURNAL of MATHEMATICAL MODELS AND METHODS IN APPLIED SCIENCES",
  year="2013",
  volume="7",
  number="8",
  pages="755--763",
  issn="1998-0140"
}