Arnaud Doucet's Suggested Papers


My schedule:
In Term 1, I can supervise a maximum of 2 papers.
In Term 2, I can supervise a maximum of 1 paper, possibly 2 (tentative).
Please choose a paper from the following list. Take a look at the paper first to see if it would be interested. Let me know if you would like to work on that paper, and I will mark it as unavailable. Contact me (in person or by e-mail) to schedule a meeting so that we can discuss exactly what you would like to do with the paper.

Stochastic computation papers:
  • Chan, C., Kohn, R. and Smith, M. Nonparametric Regression using Linear Combinations of Basis Functions. Statistics & Computing, 2000.
Note: This is a paper on Stochastic computational methods known as Markov chain Monte Carlo methods. This paper shows how it is possible to devise Markov Chain Monte Carlo methods to perform variable selection for nonparametric regression. It contains many examples and applications, some of which could be reproduced and extended as part of the project. If you chose this paper, you will also have the opportunity to do some programming/simulations.
  • Doucet, A., Godsill, S.J. and Andrieu, C. On Sequential Monte Carlo Sampling Methods for Bayesian Filtering. Statistics & Computing. vol. 10, no. 3, pp. 197-208, 2000. 
    AVAILABLE 

    Note: This is a paper on Stochastic computational methods known as Sequential Monte Carlo. This paper shows how it is possible to devise Monte Carlo methods to perform sequential Bayesian inference. It contains many examples and applications, some of which could be reproduced and extended as part of the project.     If you chose this paper, you will also have the opportunity to do some programming/simulations.

  • Fearnhead, P. Exact and Efficient Bayesian inference for Multiple Changepoint Problems. Preprint, 2005.
    AVAILABLE 

    • Note: This is a recent paper of Paul Fearnhead showing how it is possible to use clever recursions  to perform exact Bayesian inference for changepoint problems exactly with a reasonable computational complexity.  It contains many examples and applications, some of which could be reproduced and extended as part of the project. This project will involve some programming/simulations.

Bayesien Inference paper:
  • Höhle, M. and Held, L. Bayesian estimation of the size of a population. Preprint, 2005.
    AVAILABLE.
    Note: This paper presents an interesting Bayesian analysis of a classical problem: how to estimate the size of a population marked with serial numbers after only a sample of the serial numbers has been observed. This article is a good mix of methodology and applied work. Different Bayesian models could be proposed and evaluated theoretically and through simulations.

Return to the STAT548 home page