## nested sampling with a test

Posted in Books, pictures, Statistics, Travel, University life with tags , , , , , on December 5, 2014 by xi'an

On my way back from Warwick, I read through a couple preprints, including this statistical test for nested sampling algorithms by Johannes Buchner. As it happens, I had already read and commented it in July! However, without the slightest memory of it (sad, isn’t it?!), I focussed this time much more on the modification proposed to MultiNest than on the test itself, which is in fact a Kolmogorov-Smirnov test applied to a specific target function.

Indeed, when reading the proposed modification of Buchner, I thought of a modification to the modification that sounded more appealing. Without getting back  to defining nested sampling in detail, this algorithm follows a swarm of N particles within upper-level sets of the likelihood surface, each step requiring a new simulation above the current value of the likelihood. The remark that set me on this time was that we should exploit the fact that (N-1) particles were already available within this level set. And uniformly distributed herein. Therefore this particle cloud should be exploited as much as possible to return yet another particle distributed just as uniformly as the other ones (!). Buchner proposes an alternative to MultiNest based on a randomised version of the maximal distance to a neighbour and a ball centre picked at random (but not uniformly). But it would be just as feasible to draw a distance from the empirical cdf of the distances to the nearest neighbours or to the k-nearest neighbours. With some possible calibration of k. And somewhat more accurate, because this distribution represents the repartition of the particle within the upper-level set. Although I looked at it briefly in the [sluggish] metro from Roissy airport, I could not figure out a way to account for the additional point to be included in the (N-1) existing particles. That is, how to deform the empirical cdf of those distances to account for an additional point. Unless one included the just-removed particle, which is at the boundary of this upper-level set. (Or rather, which defines the boundary of this upper-level set.) I have no clear intuition as to whether or not this would amount to a uniform generation over the true upper-level set. But simulating from the distance distribution would remove (I think) the clustering effect mentioned by Buchner.

“Other priors can be mapped [into the uniform prior over the unit hypercube] using the inverse of the cumulative prior distribution.”

Hence another illustration of the addictive features of nested sampling! Each time I get back to this notion, a new understanding or reinterpretation comes to mind. In any case, an equally endless source of projects for Master students. (Not that I agree with the above quote, mind you!)

## BibTool on the air

Posted in Books, Linux, Travel, University life with tags , , , , , , , , , , on October 23, 2014 by xi'an

Yesterday night, just before leaving for Coventry, I realised I had about 30 versions of my “mother of all .bib” bib file, spread over directories and with broken links with the original mother file… (I mean, I always create bib files in new directories by a hard link,

    ln ~/mother.bib

but they eventually and inexplicably end up with a life of their own!) So I decided a Spring clean-up was in order and installed BibTool on my Linux machine to gather all those versions into a new encompassing all-inclusive bib reference. I did not take advantage of the many possibilities of the program, written by Gerd Neugebauer, but it certainly solved my problem: once I realised I had to set the variates

check.double = on
check.double.delete = on
pass.comments = off

all I had to do was to call

bibtool -s -i ../*/*.bib -o mother.bib
bibtool -d -i mother.bib -o mother.bib
bibtool -s -i mother.bib -o mother.bib


to merge all bib file and then to get rid of the duplicated entries in mother.bib (the -d option commented out the duplicates and the second call with -s removed them). And to remove the duplicated definitions in the preamble of the file. This took me very little time in the RER train from Paris-Dauphine (where I taught this morning, having a hard time to make the students envision the empirical cdf as an average of Dirac masses!) to Roissy airport, in contrast with my pedestrian replacement of all stray siblings of the mother bib into new proper hard links, one by one. I am sure there is a bash command that could have done it in one line, but I spent instead my flight to Birmingham switching all existing bib files, one by one…

## control functionals for Monte Carlo integration

Posted in Books, Statistics, University life with tags , , , , , on October 21, 2014 by xi'an

This new arXival by Chris Oates, Mark Girolami, and Nicolas Chopin (warning: they all are colleagues & friends of mine!, at least until they read those comments…) is a variation on control variates, but with a surprising twist namely that the inclusion of a control variate functional may produce a sub-root-n (i.e., faster than √n) convergence rate in the resulting estimator. Surprising as I did not know one could get to sub-root-n rates..! Now I had forgotten that Anne Philippe and I used the score in an earlier paper of ours, as a control variate for Riemann sum approximations, with faster convergence rates, but this is indeed a new twist, in particular because it produces an unbiased estimator.

The control variate writes

$\psi_\phi (x) = \nabla_x \cdot \phi(x) + \phi(x)\cdot \nabla \pi(x)$

where π is the target density and φ is a free function to be optimised. (Under the constraint that πφ is integrable. Then the expectation of ψφ is indeed zero.) The “explanation” for the sub-root-n behaviour is that ψφ is chosen as an L2 regression. When looking at the sub-root-n convergence proof, the explanation is more of a Rao-Blackwellisation type, assuming a first level convergent (or presistent) approximation to the integrand [of the above form ψφ can be found. The optimal φ is the solution of a differential equation that needs estimating and the paper concentrates on approximating strategies. This connects with Antonietta Mira’s zero variance control variates, but in a non-parametric manner, adopting a Gaussian process as the prior on the unknown φ. And this is where the huge innovation in the paper resides, I think, i.e. in assuming a Gaussian process prior on the control functional and in managing to preserve unbiasedness. As in many of its implementations, modelling by Gaussian processes offers nice features, like ψφ being itself a Gaussian process. Except that it cannot be shown to lead to presistency on a theoretical basis. Even though it appears to hold in the examples of the paper. Apart from this theoretical difficulty, the potential hardship with the method seems to be in the implementation, as there are several parameters and functionals to be calibrated, hence calling for cross-validation which may often be time-consuming. The gains are humongous, so the method should be adopted whenever the added cost in implementing it is reasonable, cost which evaluation is not clearly provided by the paper. In the toy Gaussian example where everything can be computed, I am surprised at the relatively poor performance of a Riemann sum approximation to the integral, wondering at the level of quadrature involved therein. The paper also interestingly connects with O’Hagan’s (1991) Bayes-Hermite [polynomials] quadrature and quasi-Monte Carlo [obviously!].

## a week in Warwick

Posted in Books, Kids, Running, Statistics, University life with tags , , , , , , , , , , , , on October 19, 2014 by xi'an

This past week in Warwick has been quite enjoyable and profitable, from staying once again in a math house, to taking advantage of the new bike, to having several long discussions on several prospective and exciting projects, to meeting with some of the new postdocs and visitors, to attending Tony O’Hagan’s talk on “wrong models”. And then having Simo Särkkä who was visiting Warwick this week discussing his paper with me. And Chris Oates doing the same with his recent arXival with Mark Girolami and Nicolas Chopin (soon to be commented, of course!). And managing to run in dry conditions despite the heavy rains (but in pitch dark as sunrise is now quite late, with the help of a headlamp and the beauty of a countryside starry sky). I also evaluated several students’ projects, two of which led me to wonder when using RJMCMC was appropriate in comparing two models. In addition, I also eloped one evening to visit old (1977!) friends in Northern Birmingham, despite fairly dire London Midlands performances between Coventry and Birmingham New Street, the only redeeming feature being that the connecting train there was also late by one hour! (Not mentioning the weirdest taxi-driver ever on my way back, trying to get my opinion on whether or not he should have an affair… which at least kept me awake the whole trip!) Definitely looking forward my next trip there at the end of November.

## position at Warwick

Posted in Statistics, University life with tags , , , , , , , on September 17, 2014 by xi'an

A new position for the of Professor Of Statistics and Data Science / Director of the [newly created] Warwick Data Science Institute has been posted. To quote from the job description, “the position arises from the Department of Statistics’ commitment, in collaboration with the Warwick Mathematics Institute and the Department of Computer Science, to a coherent methodological approach to the fundamentals of Data Science and the challenges of complex data sets (for example big data).”  The interview date is November 27, 2014. All details available here.

## statistical challenges in neuroscience

Posted in Books, pictures, Statistics, Travel with tags , , , , , , on September 4, 2014 by xi'an

Yet another workshop around! Still at Warwick, organised by Simon Barthelmé, Nicolas Chopin and Adam Johansen  on the theme of statistical aspects of neuroscience. Being nearby I attended a few lectures today but most talks are more topical than my current interest in the matter, plus workshop fatigue starts to appear!, and hence I will keep a low attendance for the rest of the week to take advantage of my visit here to make some progress in my research and in the preparation of the teaching semester. (Maybe paradoxically I attended a non-neuroscience talk by listening to Richard Wilkinson’s coverage of ABC methods, with an interesting stress on meta-models and the link with computer experiments. Given that we are currently re-revising our paper with Matt Moore and Kerrie Mengersen (and now Chris Drovandi), I find interesting to see a sort of convergence in our community towards a re-re-interpretation of ABC as producing an approximation of the distribution of the summary statistic itself, rather than of the original data, using auxiliary or indirect or pseudo-models like Gaussian processes. (Making the link with Mark Girolami’s talk this morning.)

## Warwick campus

Posted in pictures, Travel, University life with tags , , , , , on September 3, 2014 by xi'an