I am always interested to hear from well qualified prospective research students (particularly PhD students) who are interested in working in Computational Statistics and Probability. Below are some examples of projects I am currently interested in offering, and which might be interesting to study. Feel free to suggest other projects you may be interested in (although be aware I may not!), and contact me directly at m dot pollock at warwick dot ac dot uk.
- Bayesian Fusion (jointly with Hongsheng Dai (Essex))
- Estimating Small Probabilities (jointly with Adam Johansen)
- Inference for multi-view data, with application to criminal networks (jointly with Jim Q. Smith)
- Overview: A common statistical problem is combining inferences from multiple sources into a single coherent inference. Each of the sources may be on a common scientific problem and study the effect of the same covariates of interest, but may also study different (and only partially overlapping) covariates of interest. The difficulty with combining such inferences occurs whenever there is complicated model structure necessitating the use of computational methods to represent the inferences, and from a theoretical and methodological viewpoint would form the core of the project. This project would however consider particular application to the study of public health and criminal networks from the perspective of national security. For instance, for the application to criminal networks one may look at combing inferences which could include network analysis of criminal organisations, large data obtained for social media, criminal studies in published scientific studies, etc.
- Retrospective Monte Carlo methods (jointly with (a subset of) Paul Jenkins, Adam Johansen and Gareth Roberts)
Masters projects (details on request)
- Artificial intelligence and go
- Cracking ciphers using simulation
- Fusion applications
- Keeping a straight face
- Path-space rejection sampling for financial models
- Simulating quasi-stationary distributions
- Statistical and actuarial modelling of financial capital requirements
- Statistical topics in political opinion polling