Simultaneous Model Selection via Rate-Distortion Theory, With Applications to Cluster and Significance Analysis of Gene Expression Data

Rebecka JörnstenJournal of Computational and Graphical Statistics 2009 · DOI

Abstract

High-dimensional data are prevalent across many application areas, and generate an ever-increasing demand for statistical methods of dimension reduction, such as cluster and significance analysis. One application area that has recently received much interest is the analysis of microarray gene expression data. The results of cluster analysis are open to subjective interpretation. To facilitate the objective inference of such analyses, we use flexible parameterizations of the cluster means, paired with model selection, to generate sparse and easy-to-interpret representations of each cluster. Model selection in cluster analysis is combinatorial in the numbers of clusters and data dimensions, and thus presents a computationally challenging task. In this article we introduce a model selection method based on rate-distortion theory, which allows us to turn the combinatorial model selection problem into a fast and simultaneous selection across clusters. The method is also applicable to model selection in significance analysis We show that simultaneous model selection for cluster analysis generates objectively interpretable cluster models, and that the selection performance is competitive with a combinatorial search, at a fraction of the computational cost. Moreover, we show that the rate-distortion based significance analysis substantially increases the power compared with standard methods. This article has supplementary material online.
BibTeX
@article{jornsten2009simultaneous,
  author = {Rebecka Jörnsten},
  title = {{Simultaneous Model Selection via Rate-Distortion Theory, With Applications to Cluster and Significance Analysis of Gene Expression Data}},
  journal = {Journal of Computational and Graphical Statistics},
  year = {2009},
  doi = {10.1198/jcgs.2009.07043},
}

Loading related content...

Figure