impressions from Princeton

This one-day trip to Princeton was quite profitable thanks to the exchanges I had with the members of the econometrics department there. In particular, I really appreciated the interactive way the Gregory Chow seminar was run and the way the audience got quickly focused on the central issue of ABC, namely running inference under limited information (provided by the summary statistics). What is most interesting (to me) is that (a) the discussants focused on the limiting normal distribution as a way to bypass (in)sufficiency and (b) they did not seem to deem the use of an insufficient summary statistics a major drawback of the method. I also took advantage of this trip to correct a restricted and misguided impression on the existence of unbiased estimators, to discuss about loss functions for set estimation and empirical likelihood, and to mention the interesting paradox of the normal mean norm (Example 4.2.8 in The Bayesian Choice) where the MLE based on the distribution of the estimator of the norm improves upon this initial estimator…

 

A Bayesian Nobel recipient

Chris Sims, who got the Nobel Prize in Economics yesterday, or more precisely the Sveriges Riksbank Prize in Economic Sciences, has [also] done work in Bayesian econometrics. See for instance his talk on Why Econometrics Should Always and Everywhere Be Bayesian. Or his analysis of a counterexample of Larry Wasserman’s. He even has a tech report on adaptive Metropolis-Hastings algorithms (that apparently did not get published). He has also been teaching Bayesian statistics and econometrics at Princeton for many years, so this is a cool day for Bayesian stats! (The picture of a switching regime estimation on his webpage is actually similar to a rendering of mine from the late 90′s, when I was working on semi-Markov switching AR models with Catalin Starica, except I cannot find any trace but for the Splus code!)

European seminar on Bayesian econometrics

The European Seminar on Bayesian Econometrics will take place in Brussels, Belgium, this autumn (November 4 and 5). I wish I could have replied positively to the invitation to talk there but I already had other travel plans…

SAMSI workshop

Taking advantage of the people gathered at Frontiers of Statistical Decision Making and Bayesian Analysis, Dongchu Sun organised a one-day SAMSI workshop on reference priors for spatio-temporal models. Talking with a small group focused on this  topic was quite enjoyable and a change from the larger crowds at the conference (even though talks were also enjoyable there!). I particularly appreciated the discussion we had around AR(p) models and the difficulty of assessing whether or not non-stationary regions should be included in the analysis. The generalisation of the Berger-Yang (1994) paradigm to general values of p seems to put too much mass on the non-stationary region, even when using a symmetrisation technique… I came out of the meeting (exhausted and) wondering whether or not it was at all meaningfull to consider testing for stationarity, even though Bayes factors can be constructed in this setting.

Ritorno da Rimini

Back from the workshop in an uneventful journey where the train was on time, air-conditioned, and with a plug!, and the plane was only slightly delayed by a small storm in Bologna. The meeting was certainly interesting if very confidential (20 to 25 people) and people there quite friendly. The debate went on with no need for shouting or specious arguments, thanks to Russell’s self-confessed agnosticism. The trip also gave me the opportunity to read three short novels by Andrea Camilleri following the suggestion of Alessandra Iacobucci. Those are detective stories taking place in Sicily and I found them quite enjoyable, if not at the level of Fruttero and Lucentini’s novels. (The translation in French also misses the linguistic game between Italian, Sicilian and the intermediate versions…)

In connection with ABC in Paris, Julien Cornebise also posted on the SAMSI website a set of comments on the talks.