A hacking day of Mathematics for Data Science at the Department of Mathematics within the course of Advanced Statistical Methods

Davide Ferrari

Davide Ferrari

Free University of Bozen/Bolzano
Bio After receiving my PhD from the School of Statistics, University of Minnesota in 2008, I held tenured positions as an Assistant Professor (Ricercatore) at the University of Modena and Associate Professor (Senior Lecturer) at the University of Melbourne. I am currently an Associate Professor in the Faculty of Economics and Management at the University of Bolzano and an Associate Research Fellow at the Australian Research Council Center of Excellence for Mathematical and Statistical Frontiers (ACEMS). At the University of Bolzano, I teach theoretical and applied statistics at both undergraduate and graduate levels and I serve as Vice-Dean for teaching and director of the Masters Degree in Public Policy and Administration. I am currently interested in data integration methods, composite likelihood procedures, model selection, and inference methods for high-dimensional data. Recently, I have started to work on model selection and estimation methods for intractable likelihoods in the context of complex econometric and environmental data.


A composite likelihood (CL) is a combination of low-dimensional likelihood objects, typically corresponding to small data subsets. CL methods have received much attention in recent years due to their ability to break down complex models into simpler components and enable inference even when the full likelihood is not tractable. This short course provides students with a systematic introduction of composite likelhood modeling and inference. The course overviews the statistical theory of unbiased estimating equations, representing the foundation for composite likelihood inference. Several practical data analysis examples will be covered using the computing language R.


  • Module 1: Preliminaries and notations. Introduction to composite likelihood estimation. Examples in genetics, spatial statistics, and time-series analysis.
  • Module 2: Unbiased estimating equations, fixed-sample optimality criteria, asymptotic properties of composite likelihood estimators.
  • Module 3: Criteria for model selection, examples of composite likelihood estimation in high-dimensional settings.


Course notes and R code will be provided during the class. Selected material from the following references will be used:

  • Heyde, C.C. (2008). Quasi-likelihood and its application: a general approach to optimal parameter estimation. Springer Science and Business Media.
  • Ferguson, T.S. (1996). A Course in Large Sample Theory. London: Chapman and Hall.
  • Lindsay, B.G. (1988). Composite likelihood methods. Contemporary Mathematics,80, 220-239.
  • Lindsay, B.G., Grace Y.Y., and Sun, J. (2011) Issues and strategies in the selection of composite likelihoods. Statistica Sinica, 71-105.
  • Varin, C., Reid, N. and Firth, D. (2011). An overview of composite likelihoodmethods. Statistica Sinica, 21, 5-42.


  • Tuesday 25 May 2021 @ 16.30-18.30
  • Thursday 27 May 2021 @ 10.30-12.30
  • Friday 28 May 2021 @ 15.30-18.30


  • Participation is free, however a notification by email to Prof. Claudio Agostinelli is mandatory
  • For further information, please contact Prof. Claudio Agostinelli
  • Venue: Webinar, credentials will be sent to the participants the day before of the event
  • Language: English

Material (Restricted access, user: CL2021)