Ricordi di Belgirate

Lago Maggiore, Belgirate,
Non-parametric Bayes' parate,
Intellectual chiaccherate,
... and my, how the field has grown.
DP priors and MDPs,
Random walks on Polya trees,
Mixture models and hierarchies,
... to smooth out the wrinkles.
Exchangeability relations,
De Finetti invocations,
Markov fields on tesselations,
... but are they coherent?
Process theory flows apace,
Gaussian priors on function space,
They're popular, and have a place,
... but when did you last elicit one?
Model classes, old and new,
Dirichlet and Polya, too,
Gamma, Bernstein, Stacy (who?),
... but are they really different?
Bayesian density estimation,
The Bayesian bootstrap innovation,
And local modelling recreation,
... but what would De Finetti say?
Boats appear, and Bayesians board,
Forget the last time, priors ignored,
There'll be no mishaps ('til we're moored),
... but please pass the bottle.
Island hopping, Bayesians happy,
Unexpected hippy Frati,
Welcome to the monastery,
... but look out above!
Relaxation in the sun,
For Bayesians when the work is done,
Some stay 'til late but then must run,
... as dawn brings in the storm.
Belgirate, on the shore,
Non-parametric Bayes' galore,
And there'll be more, ancor & ancor
... in fact, we're just getting started.

Anon. 21/6/97





Return to the Conference Homepage