James: Random partition structures, Poisson Process Calculus, and Bayesian
       models
       
This talk discussses the role of random partition structures in the
posterior analysis of random probability measures and more
general random measures which may arise within the context of
Bayesian non-parametric models and spatial modelling. I will highlight
the central role of the Poisson process. As an example I  will discuss
some  connections to the Dirichlet process, Ewens sampling formula and Jim
Pitman's two-parameter Poisson-Dirichlet process and partition
distribution.