Research

Presently, I am particularly interested in a number of things:

My full publication info is here

Statistical analysis for cell biology

Modern microscopy enables cell biologist to obtain high resolution 3D images of the cell nucleus. We are especially interested in the spatial configuration of sub-nuclear bodies, and the relationship of configuration to function. This raises new challenges for imaging and analysis. Some recent work

This is on-going collaboration with Paul Freemont and Dave Stephens.

Streaming data methodology

Advances in data acquisition technology have enabled the continuous collection of data in a variety of contexts, notably sensor networks. Such contexts require data analysis algorithms that operate with a single view of each data item, and the have the capacity to adapt to changes in the underlying stochastic mechanism. Some recent work

Much of this work was partially supported by the ALADDIN project.

Classification and data mining

Classification problems are ubiquitous, and are a good example of the effectiveness of Statistics in the real world. I am especially interested in complicated contexts, like credit card fraud detection. Outputs from the EPSRC ThinkCrime project include

Data mining involves various types of analysis on large data sets. We have been involved in defining the sub-area of "pattern detection and discovery", concerned with finding small local structures in large data sets. Some work in this area includes

And a recent review-type article on data mining:

Cyber-security

Cyber-security is concerned handling the problem of malicious use of computers and networks. I am especially interested in the statistical problem of anomaly detection in network traffic data. A recent paper: We organised a workshop on network analysis in 2013. This resulted in a forthcoming edited volume.

Credit Scoring

I have a long term interest in Credit Scoring, particularly for use in classification problems. Some recent work: