Skip to main content

Dr Michael Pitt

Google scholar entry:

Recent publications:

  • Discussion: Sequential Quasi-Monte-Carlo Sampling (by Mathieu Gerber and Nicolas Chopin), lead discussion,

Journal of the Royal Statistical Society, Series B, forthcoming 2015.

Biometrika, forthcoming, 2015.

A recent talk on this (Isaac Newton Institute for Mathematic Sciences, University of Cambridge, 2014) here (to "err" is human) and slides here.

Statistics and Computing, forthcoming.

 Journal of Econometrics 179 (2), pp 99--111, May 2014.

Annals of the Institue of Statistical Mathematics, 2014.

 Journal of Computational and Graphical Statistics, 2014.

 Journal of Econometrics 171 (2), pp 134–151, Dec 2012.

Journal of Econometrics 165 (2), pp 190–209, Dec 2011.

  • Book chapter: “Bayesian Inference for Time Series State Space Models”, (with P Giordani and R Kohn), pp 71-- 124,

The Oxford Handbook of Bayesian Econometrics (OUP). Eds Geweke, J. and Koop, G. and van Dijk, H, 2011.

  • Discussion of “Particle learning for sequential Bayesian computation” by Lopes, H.F. and Carvalho, C.M. and Johannes, M. and Polson, N.G.,

Bayesian Statistics 9, 2010.

  • Discussion of “Particle Markov chain Monte Marlo methods” by Andrieu, C. and Doucet, A. and Holenstein, R (with R S Silva, R Kohn, P Giordani),

Journal of the Royal Statistical Society, Series B, 2010.

Earlier publications:

  • "Efficient Bayesian inference for Gaussian copula regression models" (with D Chan and R Kohn).

Biometrika 93(3), pages 537-554, 2006.

  • "Extended Constructions of Stationary Autoregressive Processes" (with S Walker).

Statistics and Probability Letters, 76, 1219--1224, 2006.

  • "Constructing Stationary Time Series Models using Auxiliary Variables with Applications" (with S Walker).

Journal of the American Statistical Association Vol. 100, No. 470, June 2005 pp.554-564.

  • "Correction: Likelihood analysis of non-Gaussian measurement time series" (with N. Shephard).

Biometrika, 2003.

  • "Constructing first order stationary autoregressive models via latent processes" (with C Chatfield and S Walker).

Scandinavian Journal of Statistics, 29, 657–663, 2002.

  • Comment (with S Walker) on “Non-Gaussian OU based models and some of their uses in financial economics” by O. E. Barndorff-Nielsen and N. Shephard.

Journal of the Royal Statistical Society (Series B) , 63, 2001.

  • Comment (with N Shephard) on “Time series analysis of non-Gaussian observations based on state space models from both classical and Bayesian perspectives” by J. Durbin and S. J. Koopman.

Journal of the Royal Statistical Society (Series B), 62, 2000.

  • "Analytic convergence rates and parameterisation issues for the Gibbs sampler applied to state space models" (with N. Shephard),

Journal of Time Series Analysis, 20, 63-85, 1999.

  • "Filtering via simulation: auxiliary particle filter" (with N. Shephard).

Journal of the American Statistical Association 1999, 94, 590-9.

  • "Trade Union Decline and the Distribution of Wages in the UK: Evidence from Kernel Density Estimation" (with B D Bell),

Oxford Bulletin of Economics and Statistics, Vol 60, No 4, 1998.

  • Proceedings: "A Comparison of Sample Based Filters and the Extended Kalman Filter for the Bearings-only Tracking Problem" (with N Gordon). p 2017.

Proceedings of Ninth European Signal Processing Conference, 1998.

  • "Likelihood analysis of non-Gaussian measurement time series" (with N Shephard),

Biometrika 84 , 653-67,1997.

Reprinted in "Readings in Unobserved Component Models" A.C. Harvey and T. Proietti, 2005, 368-385, Oxford University Press.

  • Proceedings: "Antithetic MCMC for non-Gaussian measurements with applications to stochastic volatility", (with N. Shephard).

Proceedings of the American Statistical Association, Bayesian Statistics Section, 1997, 81-6.

THESIS (STARTED 1995 submitted May 1997)

On Bayesian inference for non-Gaussian state space models (chapter 3 is joint work). Only first 5 chapters (next 4 are not interesting). Just MCMC - no particle filtering here.

chapter1 Introduction. Kalman filtering, smoothing. MCMC etc.

chapter2 In this chapter I use single site (specifically rejection methods) for updating states (example SV models). This is the first time this method was introduced. The method is extremely fast, robust and has acceptance probabilities of well over 99%.

chapter3 Follows on from prev chapter. The chapter which is joint work with Neil Shephard. We use blocking methods for states. This was the first time such proposals had been used.

chapter4 Shows how reparameterisation leads to quite radical differences in performance of MCMC (depending on signal to noise ratio). Also I show the convergence rate for single move methods.

chapter5 Introduces a way of joint sampling states and parameters in an MCMC scheme. A Laplace approximation to the conditional density of the parameters in the approximating SSF is introduced (the first time this had been done). Also introduces a way of importance sampling on both the states and parameters - this is not attempted as importance sampling will not work in this joint space.


 "Auxiliary variable based particle filters" (with N Shephard), in A. Doucet, J.F.G. de Freitas and N.J. Gordon (eds.). Sequential Monte Carlo Methods in Practice (New York: Springer-Verlag, 2001), 271--293 (with N Shephard).

Time varying covariances: a factor stochastic volatility approach, (with discussion) in J.M. Bernardo, J.O. Berger, A.P. Dawid and A.F.M. Smith (eds.), Bayesian Statistics 6, Proceedings of the Sixth Valencia International Meeting (Oxford: Oxford University Press, 1999), 547-570 (with N Shephard).


Working papers:

Smooth particle filters for likelihood evaluation and maximisation. To be revised and resubmitted. This version 16/07/2002.

Efficient likelihood based inference for observed and partially observed diffusions (with S Chib and N Shephard). To be submitted. This version 04/07/04.