Advertisement

Robust Estimation of Skew-Normal Parameters with Application to Outlier Labelling

  • Mario RomanazziEmail author
Conference paper
Part of the Springer Proceedings in Mathematics & Statistics book series (PROMS, volume 274)

Abstract

We suggest to estimate the parameters of the skew-normal distribution by the method of moments, modified so as to achieve robustness. A type of trimmed estimator is used, with the trimming fraction depending on a given scaled deviation from the center. An application to outlier labelling is illustrated.

Keywords

Trimmed mean \(\delta \)-trimmed moments Quantile-based estimator Skew-t Boxplot 

References

  1. 1.
    Arellano-Valle, R.B., Azzalini, A.: The centred parameterization and related quantities of the skew- t distribution. J. Multivar. Anal. 113, 73–90 (2013)MathSciNetCrossRefGoogle Scholar
  2. 2.
    Azzalini A (2014) The skew-normal and skew-t distributions. R package version 1.1-1Google Scholar
  3. 3.
    Davies, L., Gather, U.: The identification of multiple outliers. J. Am. Stat. Assoc. 88, 782–792 (1993)MathSciNetCrossRefGoogle Scholar
  4. 4.
    Hinkley, D.V.: On power transformations to symmetry. Biometrika 62, 101–111 (1975)MathSciNetCrossRefGoogle Scholar
  5. 5.
    Moors, J.J.A.: A quantile alternative for kurtosis. The Statistician 37, 25–32 (1988)CrossRefGoogle Scholar
  6. 6.
    Serfling, R.: Approximation Theorems of Mathematical Statistics. Wiley, New York (1980)CrossRefGoogle Scholar
  7. 7.
    Sim, C.H., Gan, F.F., Chang, T.C.: Outlier labeling with boxplot procedures. J. Am. Stat. Assoc. 100, 642–652 (2005)MathSciNetCrossRefGoogle Scholar
  8. 8.
    Wu, M., Zuo, Y.: Trimmed and winsorized means based on a scaled deviation. J. Stat. Plan. Inference 139, 350–365 (2009)MathSciNetCrossRefGoogle Scholar
  9. 9.
    Zuo, Y.: Multidimensional trimming based on projection depth. Ann. Stat. 34, 2211–2251 (2006)MathSciNetCrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Department of Environment, Informatics and StatisticsCa’ Foscari UniversityVeniceItaly

Personalised recommendations