Determining optimal rank in reduced rank regression model by the likelihood ratio test

Document Type : Research Paper

Author
Saad Abed Madhi
Department of Medical Instrumentation Technologies Engineering, Hilla University College, Babylon, Iraq
10.22075/ijnaa.2022.28133.3812
Abstract
This paper presents a method for determining the actual rank of the coefficients matrix in the reduced rank multivariate regression model. The method is constructed using the singular value decomposition and the Likelihood Ratio Test(LRT). Some illustrative examples are given to verify this method.
Keywords
Multivariate Reduced Rank Regression
Singular Value Decomposition
More-Penrose inverse
Maximum likelihood estimation
[1] Willin Alvarez and V. J. Griffin, Estimation procedure for reduced rank regression, passed, Statistics, Optimization and Information Computing 4 (2016), no. 2, 107–117.
[2] P. T. Davies and M. K-S. Tso, Procedures for reduced-rank regression, Applied Statistics 31 (1982), 244.
[3] Irving John Good, Some applications of the singular decomposition of a matrix, Technometrics 11 (1969), 823–831.
[4] Paul E Green, Mathematical tools for applied multivariate analysis, Academic Press, 2014.
[5] Alan Julian Izenman, Reduced-rank regression for the multivariate linear model, Journal of multivariate analysis 5 (1975), 248–264.
[6] Saad Abed Madhi, Reduced-rank regression model estimation and rank determination, University of Manchester, 1981.
[7] Saad Abed Madhi and Samira Faisal Abushilah, Selecting optimal rank in reduced rank regression by aic, IJAAR 5 (2021), 43–51.
International Journal of Nonlinear Analysis and Applications
Volume 13, Issue 2
July 2022
Pages 3173-3181

  • Receive Date 18 January 2022
  • Revise Date 05 March 2022
  • Accept Date 11 April 2022

Article View 44,182
PDF Download 50,774