Premium
Comparing estimation of the parameters of distribution of the root density of plants in the presence of outliers
Author(s) -
Jabbari Nooghabi Mehdi
Publication year - 2021
Publication title -
environmetrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.68
H-Index - 58
eISSN - 1099-095X
pISSN - 1180-4009
DOI - 10.1002/env.2676
Subject(s) - mathematics , outlier , estimator , statistics , moment (physics) , exponential distribution , percentile , exponential function , distribution (mathematics) , least squares function approximation , mathematical analysis , physics , classical mechanics
The root density of plants with depth follows exponential or the Lindley distribution in the presence of outliers generated from a uniform distribution. In this article, we estimate the parameters of the Lindley distribution in the presence of outliers generated from a uniform distribution based on the moment, maximum likelihood, least squares, weighted least squares, percentile, Cramer–von‐Mises, and Anderson–Darling methods and mixture estimator of moment and maximum likelihood. These methods of estimation are compared. Also, the estimators of the parameters of Lindley‐uniform contaminated distribution are compared with the corresponding estimators of exponential‐uniform contaminated distribution, which was presented by Dixit and Nasiri, Metron , 59(3–4), 187–198 (2001). Furthermore, an analysis of an actual example of the root length of plants is presented for illustrative purposes. It is concluded that the Lindley‐uniform contaminated distribution is more appropriate than the exponential‐uniform contaminated distribution to model the root density of plants.