UPM Institutional Repository

Computation of extreme-value parameters and inference by approximation covariance technique.


Citation

Karmokar, Provash Kumar and Shitan, Mahendran (2012) Computation of extreme-value parameters and inference by approximation covariance technique. Pakistan Journal of Statistics, 28 (2). pp. 259-269. ISSN 1012-9367

Abstract

Ordinary least squares (OLS) and Linear (LIN) estimators are commonly used in estimating the parameters of location-scale family of distributions. Various works have been done to compare the efficiency between these two estimators for the two-parameter exponential distribution and the two-parameter Weibull distribution. Motivated by these works, it would be of interest to evaluate the performance of the LIN method for the extreme-value distribution. We found that the performance of LIN estimator is better than that of OLS estimator in the sense that it had smaller standard errors and better efficiency.


Download File

[img]
Preview
PDF (Abstract)
Computation of extreme.pdf

Download (83kB) | Preview

Additional Metadata

Item Type: Article
Divisions: Institute for Mathematical Research
Faculty of Science
Keywords: Ordinary least squares; Linear estimator; Relative efficiency; Location-scale family; Generalized variance; Approximation covariance.
Depositing User: Nur Farahin Ramli
Date Deposited: 03 Oct 2013 03:49
Last Modified: 20 Sep 2015 23:57
URI: http://psasir.upm.edu.my/id/eprint/24263
Statistic Details: View Download Statistic

Actions (login required)

View Item View Item