Davide Ferrari,
Professor of Statistics
Contact
University of Bolzano
Piazza Università 1 39100,
Bolzano, Italy
Phone: +39 0471013307
E-mail: davferrari@unibz.it
About me
I am a Full professor of Statistics in the Faculty of Economics and Management at the University of Bolzano in Italy. At the University of Bolzano I teach theoretical and applied statistics at both undergraduate and graduate levels. I am currently the director of the MSc in Data Analytics for Economics and Management.
​
My research interests include data integration, composite likelihood procedures, model selection, and inference methods for high-dimensional data. Recently, I have started working on model selection and estimation methods for intractable likelihoods in the context of complex econometric and environmental data which allows me to combine my interest in likelihood-based methods with my interest in applications.
ACADEMIC POSITIONS
EDUCATION
2003 - 2008
PhD in Statistics, University of Minnesota
2003 - 2005
MSc in Statistics, University of Minnesota
1998 - 2002
Laurea in Economics, University of Modena
2016 - 2018
Senior Lecturer (US Assoc. Prof), School of Mathematics and Statistics, University of Melbourne
2012 - 2016
Lecturer (US Assist. Prof), School of Mathematics and Statistics, University of Melbourne
2009 - 2011
Ricercatore (US Assist. Prof.), University of Modena and Reggio Emilia
2021 - Present
Associate Professor, Faculty of Economics and Management, University of Bolzano
2021 - Present
Full Professor, Faculty of Economics and Management, University of Bolzano
MY LATEST TEACHING
Statistics 27010 · SECS-S/01 · Bachelor of Economics & Mgmt (L-18)
Statistics for Tourism Sports and Events 30171 · SECS-S/01 · Bachelor in Tourism, Sport and Event Mgmt (L-18)
RECENT RESEARCH
Huang, Z. and Ferrari, D, Fast construction of efficient estimating equations. The increasing size and complexity of modern data challenges the applicability of traditional likelihood-based inference. Composite likelihood (CL) methods address the difficulties related to model selection and computational intractability of the full likelihood by combining a number of low-dimensional likelihood objects into a single objective function used for inference. We introduce a procedure to combine partial likelihood objects from a large set of feasible candidates and simultaneously carry out parameter estimation. The new method constructs estimating equations balancing statistical efficiency and computing cost by minimizing an approximate distance from the full likelihood score subject to a L1-norm penalty representing the available computing resources. An asymptotic theory within a framework where both sample size and data dimension grow is developed and finite-sample properties are illustrated through numerical examples.
Zheng, C, Yang, Y. and Ferrari, D. Model selection confidence sets by likelihood ratio testing. The traditional activity of model selection aims at discovering a single model superior to other candidate models. In the presence of pronounced noise, however, multiple models are often found to explain the same data equally well. To resolve this model selection ambiguity, we introduce the general approach of model selection confidence sets (MSCSs) based on likelihood ratio testing. A MSCS is defined as a list of models statistically indistinguishable from the true model at a user-specified level of confidence, which extends the familiar notion of confidence intervals to the model-selection framework. Our approach guarantees asymptotically correct coverage probability of the true model when both sample size and model dimension increase. We derive conditions under which the MSCS contains all the relevant information about the true model structure. In addition, we propose natural statistics based on the MSCS to measure importance of variables in a principled way that accounts for the overall model uncertainty.
Illustration of the trade-off between statistical and computational efficiency for a m-variate normal random vector X. Top row: Sub-likelihood weights for increasing penalty values (lambda) with corresponding number of sub-likelihoods reported at the top. Bottom row: Asymptotic relative efficiency of the implied CL estimator compared to maximum likelihood. The columns correspond to dimension (m) and pairwise correlation (\rho) between elements in X.
Di Lascio M, Falchetta, G. and Ferrari, D., Change detection from high-resolution airborne laser scans using penalized composite likelihood screening. The study of elevation change using pre- and post-event raster grids obtained by airborne laser scans is central to many scientific areas, including geomorphology, landscape science and environmental science. Separating real elevation change from noise is a nontrivial task due to complex sources of measurement uncertainty causing errors to be spatially variable. For high-resolution grids, an additional hurdle is the excessive computational effort required by common statistical models for the spatial error. This paper introduces a screening methodology for change detection in high-resolution elevation difference grids using a composite likelihood regularization framework. We propose to model the spatial error by a sparse spatial autoregressive conditional heteroscedasticity process and obtain simultaneous detection of change location and change size estimation by maximizing a penalized composite likelihood objective. The methodology is implemented through a fast coordinate-wise maximization algorithm with computational complexity growing linearly with the grid size. Numerical studies on simulated and real data show satisfactory accuracy for both change detection and parameter estimation.