Archive for quasi-Monte Carlo methods

MCQMC2014 in Belgium

Posted in Books, Statistics, Travel, University life with tags , , , , , , , , , on October 11, 2013 by xi'an

The conference MCQMC2014 (which stands for Monte Carlo and Quasi-Monte Carlo) will take place in Leuven, Belgium, on April 6-11. More exactly, in the Katholieke Universiteit Leuven, which is the Flemish-speaking side of the split (1968) Catholic University of Leuven, the French speaking side Université Catholique de Louvain being located in Louvain-la-Neuve. After missing MCQMC2012 in Sydney,  I will attend this conference where I give an invited talk (on ABC, what else..?!). As it happens (and kind of logically), I have visited Louvain-la-Neuve many times, especially in the previous era where historical Bayesians Michel Mouchard, Jean-Marie Rolin and Léopold Simar were together in the Statistics department—two of them contributing to the highly formalised “Elements of Bayesian Statistics” that I perused during my PhD thesis in Rouen—, but I have never been to Leuven or to KU Leuven,

A great item of news is that one of the two tutorials (on April 6, 2014) will given by Art Owen, the theme being “ANOVA, global sensitivity, Sobol’ indices and all that“, The second tutorial is by Mike Giles (Oxford) on his approach of multi-level Monte Carlo methods. (If the organisers follow the MCQMC2012 trend, the Sunday afternoon tutorials should follow one another.)

9th IMACS seminar on Monte Carlo Methods, Annecy

Posted in Mountains, pictures, R, Running, Statistics, Travel, University life with tags , , , , , , , , , , on July 18, 2013 by xi'an

IMG_0361As astute ‘Og’s readers may have gathered (!), I am now in Annecy, Savoie, for the 9th IMACS seminar on Monte Carlo Methods. Where I was kindly invited to give a talk on ABC. IMACS stands for “International Association for Mathematics and Computers in Simulation” and the conference gathers themes and sensibilities I am not familiar with. And very few statisticians. For instance, I attended a stochastic particle session that had nothing to do with my understanding of particle systems (except for Pierre Del Moral’s mean field talk). The overall focus seems to stand much more around SDEs and quasi-Monte Carlo methods. Both items for which I have a genuine interest but little background, so I cannot report much on the talks I have attended beyond reporting their title. I for instance discovered the multilevel Monte Carlo techniques for SDEs, which sounds like a control variate methodology to reduce the variance w/o reducing the discretisation step. (Another instance is that the proceedings will be published in Mathematics and Computers in Simulation or Monte Carlo Methods and Applications. Two journals I have never published in.) Although I have yet a few months before attending my first MCQMC conference, I presume this is somehow a similar spirit and mix of communities.

IMG_0392At another level, attending a conference in Annecy is a blessing: the city is beautiful, the lake pristine and tantalising in the hot weather, and the surrounding mountains (we are actually quite close to Chamonix!) induce me to go running on both mornings and evenings.

Winter workshop, Gainesville (day 2)

Posted in pictures, Running, Travel, University life, Wines with tags , , , , , , , , , , , , , on January 21, 2013 by xi'an

view from the Emerson Alumni Hall auditorium,  UF, Gainesville, Florida, Jan. 19, 2013On day #2, besides my talk on “empirical Bayes” (ABCel) computation (mostly recycled from Varanasi, photos included), Christophe Andrieu gave a talk on exact approximations, using unbiased estimators of the likelihood and characterising estimators garanteeing geometric convergence (bounded weights, essentially, which is a condition popping out again and again in the Monte Carlo literature). Then Art Owen (father of empirical likelihood among other things!) spoke about QMC for MCMC, a topic that always intringued me.

Indeed, while I see the point of using QMC for specific integration problems, I am more uncertain about its relevance for statistics as a simulation device. Having points uniformly distributed over the unit hypercube in a much more efficient way than a random sample is not helping much when only a tiny region of the unit hypercube, namely the one where the likelihood concentrates, matters. (In other words, we are rarely interested in the uniform distribution over the unit hypercube: we instead want to simulate from a highly irregular and definitely concentrated distribution.) I have the same reservation about the applicability of stratified sampling: the strata have to be constructed in relation with the target distribution. The method Art advocates using a CUD (completely uniformly distributed) sequence as the underlying (deterministic) pseudo-unifom sequence. Highly interesting and I want to read the paper in greater details, but the fact that most simulation steps use a random number of uniforms seems detrimental to the performances of the method in general.

alligator in Lake Alice, Gainesville, Florida, Jan. 19, 2013After a lunch break at a terrific BBQ place, with a stop at Lake Alice to watch the alligator(s) I had missed during my morning run, I was able this time to attend till the end Xiao-Li Meng’s talk, where he presented new improvements on bridge sampling based on location-scale (or warping) transforms of the original two-samples to make them share mean and variance. Hani Doss concluded the meeting with a talk on the computation of Bayes factors when using (non-parametric) Dirichlet mixture priors, whose resolution does not require simulations for each value of the scale parameter of the Dirichlet prior, thanks to a Radon-Nykodim derivative representation. (Which nicely connected with Art’s talk in that the latter mentioned therein that most simulation methods are actually based on Riemann integration rather than Lebesgue integration. Hani’s representation is not, with nested sampling being another example.)

We ended up the day with a(nother) barbecue outside, under the stars, in the peace and quiet of a local wood, with wine and laughs, just like George would have concluded the workshop. This was a fitting ending to a meeting dedicated to his memory…

A survey of [the 60's] Monte Carlo methods

Posted in Books, R, Statistics, University life with tags , , , , , , , on May 17, 2011 by xi'an

“The only good Monte Carlos are the dead Monte Carlos” (Trotter and Tukey, quoted by Halton)

When I presented my [partial] history of MCM methods in Bristol two months ago, at the Julian Besag memorial, Christophe Andrieu mentioned a 1970 SIAM survey by John Halton on A retrospective and prospective survey of the Monte Carlo method. This is a huge paper (62 pages, 251 references) and it brings a useful light on the advances in the 60’s (the paper was written in 1968). From the reference list, it seems John Halton was planning two books on the Monte Carlo method, but a search on google did not show anything. I also discovered in this list that there was a 1954 RSS symposium (Read Paper?) on Monte Carlo methods. Note that there were at least two books on Monte Carlo published at the time, Hammersley and Handscomb’s 1964 Monte Carlo Methods and Scheider’s 1966 Monte Carlo Method. (Hammerlsey appears as a major figure in this survey.) There is a lot of material in this long review and most of the standard methods are listed: control variate, importance sampling, self-normalised simportance sampling, stratified sampling, antithetic variables, simulation by inversion, rejection or demarginalisation. Variance reduction is presented as the motivation for the alternative methods. Very little is said about the use of Monte Carlo methods in statistics (“many of  [the applications] are primitive and artless“)  I was first very surprised to find sequential Monte Carlomentioned as well, but it later appeared this was Monte Carlo methods for sequential problems, in the spirit of Abraham Wald. While the now forgotten EZH method is mentioned as a promising new method (p.11), the survey also contains an introduction to the conditional Monte Carlo method of Trotter and Tukey (1956) [from whom the above and rather puzzling quote is taken] that could relate to the averaging techniques of Kong, McCullagh, Meng, Nicolae and Tan as found in their 2003 Read Paper….

“The search for randomness is evidently futile” (Halton)

A large part of the review is taken by the study of uniform random generators and by the distinction between random, pseudo-random and quasi-random versions. Halton insists very much on the lack of justification in using non-random generators, even though they work well. He even goes further as to warn about bias because even the truly random generators are discrete. The book covers the pseudo-random generators, starting with the original version of von Neumann, Metropolis, and Ulam, continuing with Lehmer’s well-known congruencial generator, and the Fibonacci generalisation. For testing those generators by statistical tests (again with little theoretical ground), Marsaglia is mentioned.  The paper also covers in great detail the quasi-random sequences, covering low discrepancy requirements, van der Corput’s, Halton’s and Hammersley’s sequences. Halton considers quasi-Monte Carlo as “a branch of numerical analysis”.

The paper concludes with a list of 24 future developments I will cover in another post tomorrow…

MCQMC 2012

Posted in Statistics with tags , , , , , on January 26, 2011 by xi'an

Just received today the announcement for the next MCQMC meeting, which will be the 10th International Conference on Monte Carlo and Quasi-Monte Carlo Methods in Scientific Computing (MCQMC 2012), taking place in Sydney from February 13 to 17, 2012.

MCQMC is a biennial conference devoted to Monte Carlo and quasi-Monte Carlo methods and their interactions and applications.  (In brief, quasi-Monte Carlo methods replace the random choices that characterize the Monte Carlo method by well chosen deterministic choices.)  For more information, click on the “Background” tab on the web site.  This will be the first MCQMC conference to be held in the southern hemisphere.  (Northerners may like to be reminded that February is summertime in Sydney!)

The invited plenary speakers for MCQMC are
* Pierre Del Moral (INRIA & University of Bordeaux 1, France)
* Mike Giles (University of Oxford, UK)
* Fred J. Hickernell (Illinois Institute of Technology, USA)
* Aicke Hinrichs (University of Jena, Germany)
* Michael Lacey (Georgia Institute of Technology, USA)
* Kerrie Mengersen (Queensland University of Technology, Australia)
* Andreas Neuenkirch (University of Kaiserslautern, Germany)
* Art B. Owen (Stanford University, USA)
* Leszek Plaskota (University of Warsaw, Poland)
* Eckhard Platen (University of Technology Sydney, Australia)

Proposals for special sessions should contain
* a short description of the theme and scope of the session
* the name(s) of the organizer(s)
* the names of four speakers (preferably from different institutions)
who have agreed to participate
The deadline for submitting proposals is July 1, 2011.  There will be a later call for contributed talks.

Please join us for MCQMC

Follow

Get every new post delivered to your Inbox.

Join 604 other followers