Archive for the pictures Category

AISTATS 2014 [no day #2]

Posted in Kids, Mountains, pictures, Running, Travel with tags , , , on April 25, 2014 by xi'an


Even though the talks yesterday morning at AISTATS 2014 (day #2) were of the same high quality as the other days, I “had” to skip them to drive my daughter to the airport. Originally, the flight was planned at 7:40, which would have given me plenty of time to get back for Andrew’s talk at 9pm, but the airport staff was on strike from 4 to 9 and so all flights were delayed. (Compared with strikes in France which are synonyms with chaotic mess, this Icelandic strike was quite civilized in that the disruption was rather minimal, with desks opening at 9pm sharp [and the passengers in the queue counting down the last seconds together!] and a very swift processing of bags.) Once at the airport, my wife and I took the opportunity to go birding on the peninsula, unsuccessfully  seeking the elusive puffins who have seemingly not yet made it back to the Icelandic cliffs where they nest in the summer. And I managed to get back to the conference hotel for an AISTATS executive meeting which may have landed me in more involvement in the AISTATS 2016 edition of the conferences.

Then, rather than joining the whole conference happily soaking in the hot waters of the blue lagoon, we went for a hike to a peak just north of Reykjavik, Þverfellshorn, which is part of the Esjan range,starting at 4pm (a first for me as I usually prefer starting at 4am..!, but this is far North, where days are getting long enough to enjoy two days in one!), and aiming at the leftmost part of the plateau in the picture below. The path was very well-marked, with a grading in boots (from one to three), and snow-free until the ultimate slope (it would have been very different a few days ago), and very mild in difficulty. What I liked most was the great views, starting under blue skies and ending up with a stormy one, the sun being low enough to dramatically light up the scene. (An helicopter visibly looking for someone in the mountain range obviously added to the dramatic feeling.) Another nice thing was to meet many people going up on our way down, meaning that locals did see the hike as an evening stroll they could start at 7pm!


skyndimynd frá Íslandi (#3)

Posted in Kids, Mountains, pictures, Travel with tags , , , , , on April 24, 2014 by xi'an


Sumardagurinn fyrsti

Posted in Mountains, pictures, Travel with tags , , , , , on April 24, 2014 by xi'an


controlled thermodynamic integral for Bayesian model comparison

Posted in Books, pictures, Running, Statistics, University life with tags , , , , , , , , , , on April 24, 2014 by xi'an

Reykjavik1Chris Oates, Theodore Papamarkou, and Mark Girolami (all from the University of Warwick) just arXived a paper on a new form of thermodynamic integration for computing marginal likelihoods. (I had actually discussed this paper with the authors on a few occasions when visiting Warwick.) The other name of thermodynamic integration is path sampling (Gelman and Meng, 1998). In the current paper, the path goes from the prior to the posterior by a sequence of intermediary distributions using a power of the likelihood. While the path sampling technique is quite efficient a method, the authors propose to improve it through the recourse to control variates, in order to decrease the variance. The control variate is taken from Mira et al. (2013), namely a one-dimensional temperature-dependent transform of the score function. (Strictly speaking, this is an asymptotic control variate in that the mean is only asymptotically zero.) This control variate is then incorporated within the expectation inside the path sampling integral. Its arbitrary elements are then calibrated against the variance of the path sampling integral. Except for the temperature ladder where the authors use a standard geometric rate, as the approach does not account for Monte Carlo and quadrature errors. (The degree of the polynomials used in the control variates is also arbitrarily set.) Interestingly, the paper mixes a lot of recent advances, from the zero variance notion of Mira et al. (2013) to the manifold Metropolis-adjusted Langevin algorithm of Girolami and Calderhead (2011), uses as a base method pMCMC (Jasra et al., 2007). The examples processed in the paper are regression (where the controlled version truly has a zero variance!) and logistic regression (with the benchmarked Pima Indian dataset), with a counter-example of a PDE interestingly proposed in the discussion section. I quite agree with the authors that the method is difficult to envision in complex enough models. I also did not see mentions therein of the extra time involved in using this control variate idea.

skyndimynd frá Íslandi (#2)

Posted in Kids, Mountains, pictures, Travel with tags , , , , on April 23, 2014 by xi'an



Get every new post delivered to your Inbox.

Join 558 other followers