Bayesian transformation family selection: moving toward a transformed Gaussian universe

DOI10.1002/CJS.11261MaRDI QIDQ3463401zbMATH OpenOpenAlexFDO

Authors E. Charitidou, D. Fouskakis, I. Ntzoufras

Publication date 14 January 2016

Published in The Canadian Journal of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1312.3482

zbMATH Keywords

Bayesian model selectionposterior model probabilitiespower-priorprior compatibilitytransformation family selectionunit-information prior

Mathematics Subject Classification ID

Bayesian problems; characterization of Bayes procedures (62C10) Parametric hypothesis testing (62F03) Bayesian inference (62F15)

Abstract: The problem of transformation selection is thoroughly treated from a Bayesian perspective. Several families of transformations are considered with a view to achieving normality: the Box-Cox, the Modulus, the Yeo & Johnson and the Dual transformation. Markov chain Monte Carlo algorithms have been constructed in order to sample from the posterior distribution of the transformation parameter $lambd{a}_T$ associated with each competing family $T$. We investigate different approaches to constructing compatible prior distributions for $lambd{a}_T$ over alternative transformation families, using a unit-information power-prior approach and an alternative normal prior with approximate unit-information interpretation. Selection and discrimination between different transformation families is attained via posterior model probabilities. We demonstrate the efficiency of our approach using a variety of simulated datasets. Although there is no choice of transformation family that can be universally applied to all problems, empirical evidence suggests that some particular data structures are best treated by specific transformation families. For example, skewness is associated with the Box-Cox family while fat-tailed distributions are efficiently treated using the Modulus transformation.

Recommendations

• Objective Bayesian transformation and variable selection using default Bayes factors
• Bayesian robust transformation and variable selection: A unified approach
• scientific article; zbMATH DE number 3297777
• Variable selection of transformation models with unknown transformation function
• A new family of power transformations to improve normality or symmetry

Cites work

• A Bayesian approach to transformations to normality
• A comment on D. V. Lindley's statistical paradox
• A modified family of power transformations
• A new family of power transformations to improve normality or symmetry
• A STATISTICAL PARADOX
• An Alternative Family of Transformations
• An exploration of aspects of Bayesian multiple testing
• Bayes and empirical-Bayes multiplicity adjustment in the variable-selection problem
• Bayesian Graphical Models for Discrete Data
• Bayesian Modeling Using WinBUGS
• Bayesian robust transformation and variable selection: A unified approach
• Cluster analysis, model selection, and prior distributions on models
• Conjugate and conditional conjugate Bayesian analysis of discrete graphical models of marginal independence
• Empirical Bayes vs. fully Bayes variable selection
• scientific article; zbMATH DE number 3204183 (Why is no real title available?)
• scientific article; zbMATH DE number 3251902 (Why is no real title available?)
• Joint specification of model space and parameter space prior distributions
• Marginal Likelihood from the Gibbs Output
• Marginal Likelihood From the Metropolis–Hastings Output
• On Optimality Properties of the Power Prior
• On the Comparative Anatomy of Transformations
• Statistical quality control. A modern introduction

Cited in

(2)

• Objective Bayesian transformation and variable selection using default Bayes factors
• Joint Bayesian analysis of multiple response-types using the hierarchical generalized transformation model

Describes a project that uses

Uses Software

• WinBUGS

This page was built for publication: Bayesian transformation family selection: moving toward a transformed Gaussian universe

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3463401)