"Everything should be made as simple as possible, but not simpler." - Albert Einstein

Seminar Announcement

A Handbook for Computational Statistics Including a New Application of Reversible Jump Markov Chain Monte Carlo for Gene Expression Clustering

Ben Bird, STAT-MS Candidate, Colorado State University

Friday, December 17, 2010

9:00 a.m., room 006, Statistics Bldg

ABSTRACT

Implementations are presented in the R statistical software language for sixty examples from the textbook Computational Statistics (Givens and Hoeting, 2005). This code is intended to serve as a companion document for readers since no such resource currently exists. In addition to the code for the examples, we also provide two new examples. A new example for Reversible jump Markov chain Monte Carlo (RJMCMC) applied to cluster analysis for gene expression profiling is presented with specific application to temporal mRNA expressions of 112 genes over nine time points in rat central nervous system development. We compare this modern technique to the traditional method of k-means. Using the RJMCMC approach we build comparable clusters with greater flexibility and without the need to specify the number of groups in advance. Another new example pertains to the EM algorithm and will be briefly presented as well.

Advisory Committee:

Dr. Geof Givens, Advisor, Statistics Department, CSU

Dr. Jennifer Hoeting, Committee Member, Statistics Department, CSU

Dr. Ross Beveridge, Outside Member, Computer Science, CSU