International Workshop on Compositional Data Analysis

CoDaWork 2015: Compositional Data Analysis pp 127-143

A Regression Model for Compositional Data Based on the Shifted-Dirichlet Distribution

  • G. S. Monti
  • G. Mateu-Figueras
  • V. Pawlowsky-Glahn
  • J. J. Egozcue
Conference paper

DOI: 10.1007/978-3-319-44811-4_9

Volume 187 of the book series Springer Proceedings in Mathematics & Statistics (PROMS)
Cite this paper as:
Monti G.S., Mateu-Figueras G., Pawlowsky-Glahn V., Egozcue J.J. (2016) A Regression Model for Compositional Data Based on the Shifted-Dirichlet Distribution. In: Martín-Fernández J., Thió-Henestrosa S. (eds) Compositional Data Analysis. CoDaWork 2015. Springer Proceedings in Mathematics & Statistics, vol 187. Springer, Cham

Abstract

Using an approach based on the Aitchison geometry of the simplex, a Shifted-Dirichlet covariate model is obtained. Allowing the parameters to change linearly with a set of covariates, their effects on the relative contributions of different components in a composition are assessed. An application of this model to sedimentary petrography is given.

Keywords

Dirichlet regressionSimplicial regressionModel selection

Copyright information

© Springer International Publishing Switzerland 2016

Authors and Affiliations

  • G. S. Monti
    • 1
  • G. Mateu-Figueras
    • 2
  • V. Pawlowsky-Glahn
    • 2
  • J. J. Egozcue
    • 3
  1. 1.Department of Economics, Management and StatisticsUniversity of Milano-BicoccaMilanoItaly
  2. 2.Department of Computer Science, Applied Mathematics and StatisticsUniversity of GironaGironaSpain
  3. 3.Department of Civil and Environmental EngineeringTechnical University of CataloniaBarcelonaSpain