Archive for likelihood

workshop a Padua

Posted in pictures, Running, Statistics, Travel, University life with tags , , , , , , , , on October 5, 2012 by xi'an

I am invited to a (closed) workshop in Padua/Padova next March, “Recent advances in statistical inference: theory and case studies”, which is an exciting opportunity to discuss about recent advances in Bayesian methodology and likelihood inference, to meet with friends and to be back in this beautiful city where I met George Casella for the last time. (Keeping this vivid image of watching George running around the Prato della Valle as my bus was leaving the city towards Venezia airport.)

The workshop is organised in my favourite way, which is “to have a moderate number of invited talks at the workshop, to allow good time for presentation and discussion”. With discussants, which seems a vanishing structure in conferences where the length of the talks is getting shorter and shorter. When in Bristol last week, I realised how much I gained from a slower conference pace with fewer and longer talks, more time for discussion in between, and a well-scheduled poster session. Maybe old age speaking! Furthermore, part of the workshop takes place in the fabulous Caffè Pedrocchi, where we had dinner two years ago… Terrific (and exclusive, as the workshop is by invitation only!)

ABC à l’X

Posted in Statistics, University life with tags , , on February 7, 2011 by xi'an

Tomorrow, there is a series of seminars at École Polytechnique (X) on random models for ecology, genetics and evolution. The first one is on ABC,  Approximate Bayesian Computations Done Exactly,  by Razeesh Shainudin and I plan to attend. Here is the abstract:

Evaluating the likelihood function of parameters in highly-structured population genetic models from extant deoxyribonucleic acid (DNA) sequences is computationally prohibitive. In such cases, one may approximately infer the parameters from summary statistics of the data such as the site-frequency-spectrum (SFS) or its linear combinations. Such methods are known as approximate likelihood or Bayesian computations. Using a controlled lumped Markov chain and computational commutative algebraic methods we compute the exact likelihood of the SFS and many classical linear combinations of it at a non-recombining locus that is neutrally evolving under the infi™nitely-many-sites mutation model. Using a partially ordered graph of coalescent experiments around the SFS we provide a decision-theoretic framework for approximate sufficiency. We also extend a family of classical hypothesis tests of standard neutrality at a non-recombining locus based on the SFS to a more powerful version that conditions on the topological information provided by the SFS. Keywords: controlled lumped Markov chain, unlabelled coalescent, random integer partition sequences, partially ordered experiments, population genomic inference population genetic Markov bases, approximate Bayesian computation done exactly.