[Statlist] Research Seminar in Statistics *FRIDAY 4 MAY 2018* GSEM, University of Geneva

gsem-support-instituts g@em-@upport-|n@t|tut@ @end|ng |rom un|ge@ch
Mon Apr 30 11:13:01 CEST 2018


Dear All,

We are pleased to invite you to our next Research Seminar.

Looking forward to seeing you

Organizers :                                                                                   
E. Cantoni - S. Engelke - D. La Vecchia - E. Ronchetti
S. Sperlich - F. Trojani - M.-P. Victoria-Feser

FRIDAY 4 MAY 2018 at 11:15am, Uni-Mail M 5220

Fast construction of efficient composite estimating equations

Davide FERRARI - Free University of Bozen-Bolzano, Italy

ABSTRACT:
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. In this talk, I will discuss a new procedure to combine partial likelihood objects from a large set of feasible candidates and simultaneously carry out parameter estimation. The proposed 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. This results in truncated CL equations containing only the most informative partial likelihood score terms whilst the noisy or redundant terms are dropped. Asymptotic results within a framework where both sample size and data dimension grow are derived and finite-sample properties are illustrated through numerical examples.


Visit the website: https://www.unige.ch/gsem/en/research/seminars/rcs/




More information about the Statlist mailing list