Pseudo Likelihoods is a set of lectures within the course of Advanced Statistical Methods

for the Mathematics for daTa scieNce study plan

Helen Ogden

Helen Ogden

(University of Southampton)
Personal website
Bio Helen Ogden is a Lecturer in Statistics in the School of Mathematical Sciences at the University of Southampton, UK, and a member of the Southampton Statistical Sciences Research Institute. She received her PhD from the University of Warwick in 2014, with a thesis entitled "Inference for Generalized Linear Mixed Models with Sparse Structure", supervised by David Firth. After her PhD, she continued at Warwick as a postdoc, working on the project "Intractable Likelihoods: New Challenges from Modern Applications", before moving to Southampton in 2016. Since January 2019, she is also a Turing Fellow, as part of the Alan Turing Institute, the UK national institute for data science and artificial intelligence. Her research interests include latent variables models and methods for conducting inference when the likelihood function is difficult to compute, including composite likelihoods and numerical likelihood approximations.


  • Probability
  • Statistical Inference


These set of lectures provide an introduction to the ideas and methods for solving inferential problems in complex models. Both methods and theoretical aspectes will be covered. Illustrative examples will be provided throughout.


  • Examples of models with intractable likelihoods, with a focus on models from spatial statistics: Markov random fields, Spatially correlated random effects, Spatial extremes
  • Definition of composite likelihood, including pairwise likelihood and Besag’s pseudo-likelihood
  • Asymptotic properties of composite likelihood inference
  • Confidence intervals and hypothesis testing with composite likelihoods
  • Bayesian inference with composite likelihoods
  • Alternative approaches to inference in models with intractable likelihoods


The main reference is
- Varin, C., Reid, N., & Firth, D. (2011). An overview of composite likelihood methods. Statistica Sinica, 5-42.

For the Bayesian inference part refer to
- Pauli, F., Racugno, W., and Ventura, L. (2011). Bayesian composite marginal likelihoods. Statistica Sinica, 149-164.
- Ribatet, M., Cooley, D., & Davison, A. C. (2012). Bayesian inference from composite likelihoods, with an application to spatial extremes. Statistica Sinica, 813-845.


  • 2019/05/21 9.30-11.30, Room A213 @ Povo 1
  • 2019/05/22 9.30-11.30, Room A218 @ Povo 1
  • 2019/05/23 9.30-11.30, Room A212 @ Povo 1
  • 2019/05/24 9.30-11.30, Room A224 @ Povo 1
  • 2019/05/24 14.30-18.30, Room A223 @ Povo 1


  • Poster: PDF
  • Venue: Polo Scientifico e Tecnologico F. Ferrari
  • Language: English
  • The participation is free. Please send an email to Prof. Claudio Agostinelli.
  • For further information, please contact Prof. Claudio Agostinelli

Material (Restricted access, user: PL2018)