Archive for point processes

Latent Gaussian Models im Zürich [day 2]

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

The second day at the Latent Gaussian Models workshop in Zürich was equally interesting. Among the morning talks, let me mention Daniel Bové who gave a talk connected with the hyper-g prior paper he wrote with Leo Held (commented in an earlier post) and the duo of Janine Illian and Daniel Simpson who gave enthusiastic arguments as to why point pattern datasets should be analysed in a completely novel way, using partial SDEs. And showed us how this could be done via INLA. This perspective (purposedly?) contrasted with the modelling assumptions of Alan Gelfand who concluded the meeting with a highly interesting modelling/estimation of species distribution in the Cape area. He also ran a comparison with the Maxent approach to the same problem. As for my own talk, I somehow spent too much time on the introduction to ABC, trying to link the method with non-parametric perspectives, and so ended rushing through the sufficiency part and the population genetic results obtained by Jean-Marie Cornuet and Jean-Michel Marin the previous day. (The updated slides are available on slideshare.) I hope the main message was still spelled out clearly enough… In conclusion, this was a very interesting workshop, maybe the first of a kind since there is a possible follow-up next year in Trondheim. It showed the clear emergence of a very active INLA community, able to tackle old and new problems using this new technology, and illustrated once again the importance of developing user-friendly codes for promoting such technologies.

Seminar im Heidelberg

Posted in Books, pictures, Running, Statistics, Travel with tags , , , , on April 18, 2010 by xi'an

Even though I can report only today for unexpected family issues (and not at all for the Icelandic volcano ashes!), I have had a wonderful trip to Heidelberg! Both seminars were packed with students, many more than faculty, I met a PhD student working on ABC who travelled all the way from Bonn to attend the seminars and discuss with me, I had highly informative statistics discussions with the local faculty (about ABC, point processes, particle filters, graphical models, stochastic volatility, covariance estimation, and more), plus great runs, up the local hills, a few glimpses of the old city, and good local beer and food! Had the internet connection worked (better), I would have been a bit more efficient, but this led me to take the philosophy way for a few days. and progress in my evaliation of Evidence and Evolution. Thanks to Tilmann Gneiting for the invitation and for the warm welcome!