Visualizations for Assessing Convergence and Mixing of MCMC

  • Jarkko Venna
  • Samuel Kaski
  • Jaakko Peltonen
Conference paper

DOI: 10.1007/10955628_39

Volume 2837 of the book series Lecture Notes in Computer Science (LNCS)
Cite this paper as:
Venna J., Kaski S., Peltonen J. (2003) Visualizations for Assessing Convergence and Mixing of MCMC. In: Lavrač N., Gamberger D., Blockeel H., Todorovski L. (eds) Machine Learning: ECML 2003. ECML 2003. Lecture Notes in Computer Science, vol 2837. Springer, Berlin, Heidelberg

Abstract

Bayesian inference often requires approximating the posterior distribution with Markov Chain Monte Carlo (MCMC) sampling. A central problem with MCMC is how to detect whether the simulation has converged. The samples come from the true posterior distribution only after convergence. A common solution is to start several simulations from different starting points, and measure overlap of the different chains. We point out that Linear Discriminant Analysis (LDA) minimizes the overlap measured by the usual multivariate overlap measure. Hence, LDA is a justified method for visualizing convergence. However, LDA makes restrictive assumptions about the distributions of the chains and their relationships. These restrictions can be relaxed by a recently introduced extension.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag Berlin Heidelberg 2003

Authors and Affiliations

  • Jarkko Venna
    • 1
  • Samuel Kaski
    • 1
  • Jaakko Peltonen
    • 1
  1. 1.Neural Networks Research CentreHelsinki University of TechnologyHUTFinland