Revistes Publicacions URV: SORT - Statistics and Operations Research Transactions> 2015

A comparison of computational approaches for maximum likelihood estimation of the Dirichlet parameters on high-dimensional data

  • Dades identificatives

    Identificador: RP:2427
    Handle: https://hdl.handle.net/20.500.11797/RP2427
    Autors:
    Wehrens, RonGiordan, Marco
    Resum:
    Likelihood estimates of the Dirichlet distribution parameters can be obtained only through numerical algorithms. Such algorithms can provide estimates outside the correct range for the parameters and/or can require a large amount of iterations to reach convergence. These problems can be aggravated if good starting values are not provided. In this paper we discuss several approaches that can partially avoid these problems providing a good trade-off between efficiency and stability. The performances of these approaches are compared on high-dimensional real and simulated data.

  • Altres:

    Autor/s de la URV: Wehrens, Ron Giordan, Marco
    Paraules clau: Levenberg-Marquardt algorithm, re-parametrization, starting values, metabolomics data
    Resum: Likelihood estimates of the Dirichlet distribution parameters can be obtained only through numerical algorithms. Such algorithms can provide estimates outside the correct range for the parameters and/or can require a large amount of iterations to reach convergence. These problems can be aggravated if good starting values are not provided. In this paper we discuss several approaches that can partially avoid these problems providing a good trade-off between efficiency and stability. The performances of these approaches are compared on high-dimensional real and simulated data.
    Any de publicació de la revista: 2015
    Tipus de publicació: info:eu-repo/semantics/publishedVersion info:eu-repo/semantics/article

  • Paraules clau:

    Levenberg-Marquardt algorithm, re-parametrization, starting values, metabolomics data

  • Documents:

    DocumentPrincipal

  • Cerca a google

    Search to google scholar