Home Page of Professor Chris Holmes

Professor Chris Holmes

.....please note this page is around 4 years out of date as of 9/9/09.....I hope to find time to update it in the coming weeks.....

Room No : 40.10 Oxford Centre for Gene Function
Telephone No. : +44 (0)1865 285 368
E-Mail Address : c.holmes {a\t} stats {dot.} ox . ac .uk

Research Interests :

My main research interest lies in the field of machine learning and pattern analysis via probabalistic data modelling and the use of subjective probability theory as a unified framework for coherent inference. This has lead me to investigate Bayesian nonparametric (nonlinear) methods which include multivariate linear splines, Gaussian process priors, wavelets, neural networks, radial basis functions and local polynomial models.

My main application areas are in bioinformatics, statistical genomics and spatial statistics.

I'm based in the new Oxford Centre for Gene Function, see the link from here.

I also hold a joint appointment with the MRC Mammalian Genetics Unit at Harwell.

GPUs for statistical analysis and statistical computation:

We have been investigating the potential of graphics cards to implement highly parallel stochastic simulation algorithms (such as MCMC and SMC). We have set up a web page as a resource for researchers in statistics and related disciplines who are interested in exploring this technology. Follow the link here.

Book:

I have a monograph, co-authored with Dave Denison, Bani Mallick and Adrian Smith, that brings together current work in the area of Bayesian approaches to nonlinear data modelling. Further details can be found here.

Code From my book with Denison, Mallick and Smith:

Matlab code for implementing a variety of Bayesian nonlinear methods for classification and regression can be found in a tar file on this web site.

Conferences:

We are organising a meeting on ``Statistical Challenges Arising from Genome Resequencing'' to be held at the Isaac Newton Institute for Mathematical Sciences, Cambridge, in Summer 2010.

I co-organised a workshop on Nonlinear estimation and classification which was held at the Mathmatical Science Research Institute , Berkeley, California in the Spring of 2001. The conference brought together leading researchers in statistical data modelling from the fields of computer science, statistics and engineering. A book of the proceedings should be appearing sometime in 2002. All of the talks are available as on-line video at this page (March 19-29, 2001).

Some Publications and Technical Reports:

A more up to date list of publications is available here

Older Publications

Pintore, A., Speckman, P. and Holmes, C.C. (2004) Spatially adaptive smoothing splines. A revised version to appear in Biometrika

Pintore, A. and Holmes, C.C. (2004) Non-stationary covariance functions via spatially adaptive spectra. A revised version "tentatively accepted" in J. Amer. Statist. Assoc.

Holmes, C.C. and Held, K. (2004) Bayesian auxiliary variable models for binary and polychotomous regression. A revised version to appear in Bayesian Analysis.

Heard, N., Holmes, C.C. and Stephens, D. (submitted 2002!) A quantitative study of gene regulation involved in the immune response of Anopheline Mosquuitoes: An application of Bayesian Hierarchical Clustering of Curves. A revised version to appear in J. Amer. Statist. Assoc.

Holmes, C.C. and Brown, S. (2004) All systems GO for understanding Mouse gene function. Journal of Biology . Invited review of The functional landscape of Mouse gene expression.

Kim, H-M., Mallick, B. and Holmes, C.C. (2004) Ananlyzing nonstationary spatial data using a piecewise Gaussian process. A revised version to appear in J. Amer. Statist. Assoc.

Jasra, A., Holmes, C.C. and Stephens, D. (2004) MCMC and the label switching problem in Bayesian Mixture Models. A revised version to appear in Statistical Science.

Holmes, C.C. and Mallick, B.K. (2003) Bayesian regression with multivariate linear splines. A revised version to appear in J. Amer. Statist. Assoc.

Holmes, C.C. and Heard, N.A. (2003) Generalised monotonic regression using random change points. A revised version has appeared in Stats. in Medicine 22, pp 623--638.

Holmes, C.C. and Adams. (2003) Likelihood inference in nearest-neighbour classification models. A revised version to appear in Biometrika.

Holmes, C.C. and Adams. (2002) A probabilistic nearest-neighbour algorithm for statistical pattern recognition. A revised version has now appeared in J. Roy. Statist. Soc. Series B., vol. 64. part 2, pp 295-306.

Holmes, C.C. and Denison, D. (2002) Pefect sampling for wavelet reconstruction of signals. A revised version is found in IEEE Trans. Signal Processing. 50, pp 237--244.

Holmes, C.C. and Denison, D. (2002) Classification with Bayesian MARS. A revised version (including a couple of extra examples) to appear in Machine Learning.

Guglielmi, A., Holmes, C.C. and Walker, S.G. (2002) Perfect simulation involving a continuous and unbounded state space. A revised version to appear in J. Comp. Graph. Stats.

Holmes, C.C. and Mallick, B.K. (2001) Bayesian regression with multivariate linear splines. A revised version has now appeared in J. Roy. Statist. Soc. Series B., vol. 63, no. 1, pp 3-18.

Denison, D.G.T. and Holmes, C.C. (2001) Bayesian partitioning for estimating disease risk. A revised version has now appeared in Biometrics , vol. 57, no. 1, pp 143-149.

Roberts, S., Holmes, C.C. and Denison, D. (2001). Minimum entropy data partitioning using reversible jump Markov chain Monte Carlo. IEEE Trans. Pattern analysis and Machine Intelligence, vol. 23, June.

Holmes, C.C. and Mallick, B.K. (2000). Bayesian wavelet networks for nonparametric regression. A revised version has now appeared in IEEE. Trans. Neural Networks , vol. 11, no. 1, pp 27-35.

Holmes, C.C. and Denison, D.G.T. (1999). Bayesian wavelet analysis with a model complexity prior. A revised version has now appeared in Bayesian Statistics 6 : Oxford University Press

Holmes, C.C. and Mallick, B.K. (1998). Bayesian radial basis functions of variable dimension. Neural Computation, vol. 10, pp 1217-1233.

Technical Reports

Bayesian Co-clustering of Anopheles Gene Expression Time Series: A Study of Immune Defense Response To Multiple Experimental Challenges. (2005) (joint work with Nick Heard, Dave Stephens, George Dimopoulos and David Hand)
Bayesian model order determination and basis selection for seemingly unrelated regressions. (2001) (joint work with Bani Mallick and Dave Denison)

Bayesian partitioning for classification and regression. (2001) (joint work with Bani Mallick and Dave Denison)

Perfect simulation for Bayesian curve and surface fitting. (2001) (joint work with Bani Mallick). This paper on "coupled Markov chains from the past" is a revised version of an older paper. Perfect simulation for orthogonal model mixing.

Consultancy

I do consultancy for a mobile phone deal / price comparison engine called BillMonitor which uses statistical methods to find optimal calling plans available from the major phone operators.

Some people who I work with

Bani Mallick at Texas A&M.

Dave Denison was at Imperial College.

Steve Roberts at the robotics research group in Oxford.

Steve Walker at the University of Bath.

David Hand at Imperial College.

Back to the statistics department's home page