Robust estimation of the parameters of g-and-h distributions, with applications to outlier detection

Authors:
Yihuan Xu;Boris Iglewicz;Inna Chervoneva
Affiliations:
-;-;-
Venue:
Computational Statistics & Data Analysis
Year:
2014

Citing 4
Cited 0

Algorithms and complexity for least median of squares regression

Discrete Applied Mathematics
Computing the exact least median of squares estimate and stability diagnostics in multiple linear regression

SIAM Journal on Scientific Computing
Numerical maximum likelihood estimation for the g-and-k and generalized g-and-h distributions

Statistics and Computing
Weighted quantile-based estimation for a class of transformation distributions

Computational Statistics & Data Analysis

Quantified Score

Hi-index	0.03

Visualization

Abstract

The g-and-h distributional family is generated from a relatively simple transformation of the standard normal and can approximate a broad spectrum of distributions. Consequently, it is easy to use in simulation studies and has been applied in multiple areas, including risk management, stock return analysis and missing data imputation studies. A rapidly convergent quantile based least squares (QLS) estimation method to fit the g-and-h distributional family parameters is proposed and then extended to a robust version. The robust version is then used as a more general outlier detection approach. Several properties of the QLS method are derived and comparisons made with competing methods through simulation. Real data examples of microarray and stock index data are used as illustrations.