i-like[d the] workshop

Indeed, I liked the i-like workshop very much. Among the many interesting talks of the past two days (incl. Cristiano Varin’s ranking of Series B as the top influential stat. journal!) , Matti Vihola’s and Nicolas Chopin’s had the strongest impact on me (to the point of scribbling in my notebook). In a joint work with Christophe Andrieu, Matti focussed on evaluating the impact of replacing the target with an unbiased estimate in a Metropolis-Hastings algorithm. In particular, they found necessary and sufficient conditions for keeping geometric and uniform ergodicity. My question (asked by Iain Murray) was whether they had derived ways of selecting the number of terms in the unbiased estimator towards maximal efficiency. I also wonder if optimal reparameterisations can be found in this sense (since unbiased estimators remain unbiased after reparameterisation).

Nicolas’ talk was about particle Gibbs sampling, a joint paper with Sumeet Singh recently arXived. I did not catch the whole detail of their method but/as I got intrigued by a property of Marc Beaumont’s algorithm (the very same algorithm used by Matti & Christophe). Indeed, the notion is that an unbiased estimator of the target distribution can be found in missing variable settings by picking an importance sampling distribution q on those variables. This representation leads to a pseudo-target Metropolis-Hastings algorithm. In the stationary regime, there exists a way to derive an “exact” simulation from the joint posterior on (parameter,latent). All the remaining/rejected latents are then distributed from the proposal q. What I do not see is how this impacts the next MCMC move since it implies generating a new sample of latent variables. I spoke with Nicolas about this over breakfast: the explanation is that this re-generated set of latent variables can be used in the denominator of the Metropolis-Hastings acceptance probability and is validated as a Gibbs step. (Incidentally, it may be seen as a regeneration event as well.)

Furthermore, I had a terrific run in the rising sun (at 5am) all the way to Kenilworth where I was a deer, pheasants and plenty of rabbits. (As well as this sculpture that now appears to me as being a wee sexist…)

i-like workshop [talk]

Here are the slides of my talk at the i-like workshop in Warwick today:

I am really glad I could make it there and meet with many (highly supportive) friends for three days! The slides are quite similar to those I presented in Padova. I just added a few perspective slides…

I like…intractable likelihoods (openings)

A new EPSRC programme grant, called i-like, has been awarded to researchers in Bristol, Lancaster, Oxford, and Warwick, to conduct research on intractable likelihoods. (I am also associated to this program as a [grateful] collaborator.) This covers several areas of statistics, like big data and inference on stochastic process, but my own primary interest in the programme is of course the possibilities to conduct collaboration on ABC and composite likelihood methods. (Great website design, by the way!)

A first announcement is that there will be a half-day launch in Oxford on January 31, 2013, which program is now available. Followed by a workshop in mid-May in Warwick (to which I will participate). This event is particularly aimed at PhD students and early-career researchers. The second announcement is that the EPSRC programme grant provides funding for five postdoctoral positions over a duration of four years, which is of course stupendous! So if you like i-like as much as I like it, and are a new researcher looking for opportunities in exciting areas, you should definitely consider applying!

Structure and uncertainty, Bristol, Sept. 25

This was a fairly full day at the Structure and uncertainty modelling, inference and computation in complex stochastic systems workshop! After a good one hour run around the Clifton Down, the morning was organised around likelihood-free methods, mostly ABC, plus Arnaud Doucet’s study of methods based on unbiased estimators of the likelihood (à la Beaumont, with the novelty of assessing the inefficiency due to the estimation, really fascinating..). The afternoon was dedicated to graphical models. Nicolas Chopin gave an updated version of his Kyoto talk on EP-ABC where he resorted to composite likelihoods for hidden Markov models, (I then wondered about the parameterisation and the tolerance determination for this algorithm.) Oliver Ratman presented some of the work he did on the flu while in Duke, then move to a new approach for ABC tolerance based on various kinds of testing (which I found clearer than in Kyoto, maybe because I was not jet-lagged!) And I gave my talk on ABC-EL.I found the afternoon session harder to follow, mostly because I always have trouble understanding the motivations and the notations used on these models, albeit fascinating. I remained intrigued by the bidirectional dependence arrow in those graphs for the whole afternoon (even though I think I get it now!) After looking at the few posters presented this afternoon, I went for another short run in Leigh Woods, before joining a group of friends for an Indian dinner at the Brunel Raj. A very full day…!

semi-automatic ABC

The talk of Wednesday afternoon Ordinary Meeting of the Royal Statistical Society went on quite well, I think. I would have expected a few more people (in general) and some specific people (in particular) but this being the last week of term the schedule was not the best of times. Paul Fearnhead gave the talk, insisting on the fact that ABC had to be evaluated on its own and relating the method to Rubin‘s 1984 and to Dingle and Gratton‘ [also] 1984 papers. The main discussant was Marc Beaumont, who gave a balanced overview of the paper along with connections with population genetics. As a seconder, I managed to cover most of the points I had prepared, even though I forgot the ritual final sentence (that I was seconding the vote of thanks), as in most of the past occasions I had to discuss a Read Paper…  Most of the discussants focussed on alternative ways to implement ABC, with one particularly interesting last minute intervention about the inflation represented by associating a summary statistic with each quantity of interest. As there were many written contributions as well, I am looking forward the published collection of the discussions (I wish were available on-line).