However, candidates must first go through the national “qualification”. This process should not be problematic, but is held much earlier in the year: you need to sign up by *25 October* (next week!), then send some documents by December. Unfortunately, the committee cannot consider applications from candidates who do not hold the “qualification”.

If you need help with the process, feel free to contact me.

]]>

My first foray into Statistics was an analysis of Cox models I did for my undegraduate thesis at ENS in 2005. I had no idea back then that David Cox was still alive and active; in my mind, he was a historical figure, on par with other great mathematicians who gave their names to objects of study — Euler, Galois, Lebesgue…

When I arrived at Oxford a few months later, I was amazed to meet him, and to see that he was still very active, both as a researcher and as the organizer of events for doctoral students.

David Cox is the perfect choice as the first person to receive this prize. I hope that the inauguration of this prize will help show the public that Statistics require complex and innovative methods, that have been tackled by some exceptional minds, and should not be seen as a “sub-science” compared to other more “noble” sciences.

]]>

I am organizing a session on Wednesday morning on Advances in Monte Carlo motivated by applications; I’m looking forward to hearing the talks of Alexis Muir-Watt, Simon Barthelmé, Lawrence Murray and Rémi Bardenet during that session, as well as the rest of the very strong programme.

I’ll also be part of the jury for the best poster prize; there are many promising abstracts.

]]>

Nils Lid Hjort gave a lecture on his “confidence distributions”, a way to represent uncertainty in the non-Bayesian framework. Although he gave examples where his representation seems to work best, I wondered how this could extend to cases where the parameter is not unidimensional.

Chris Yau received the 2010 Corcoran prize and gave a short talk on applications of HMMs togenetic data; he was unlucky to have his 15-minute talk interrupted by a fire alarm (but that allowed me to wonder at how calmly efficient the British are at evacuating in such situations). Luckily, my own talk suffered no such interruption.

Peter Donnelly demonstrated once again his amazing lecturing skills, with a highly informative talk on statistical inference of the history of the UK using genetic data.

All in all, a very enjoyable afternoon, which was followed by a lovely dinner at Somerville College, with several speeches on the past, present and future of Statistics at Oxford.

Thanks again to the Corcoran committe, especially Steffen Lauritzen, for selecting me as the prize winner!

]]>

In particular, I’ll keep posting in English for anything related to my research topics.

]]>

If I had a dime for every time my wife threatened to divorce me during the past three years, I would be wealthy and not have to take a postdoctoral position which will only make me a little less poor and will keep me away from home and in the lab even more than graduate school and all because my committee read this manuscript and said that the only alternative to signing the approval to this dissertation was to give me a job mowing the grass on campus but the Physical Plant would not hire me on account of they said I was over-educated and needed to improve my dexterity skills like picking my nose while driving a tractor-mower over poor defenseless squirrels that were eating the nuts they stole from the medical students’ lunches on Tuesday afternoon following the Biochemistry quiz which they all did not pass and blamed on me because they said a tutor was supposed to come with a 30-day money-back guarantee and I am supposed to thank someone for all this?!!

(From a UMI press release, quoted in *The Whole Library Handbook 2*, 1995)

Source: Futility Closet via Arthur Charpentier.

]]>

A few months ago, I might have written

$\lambda/\mu=\kappa/\nu \Rightarrow \exists \Theta,\forall i, \sum_{j\in\mathbb{N}} E[D_{i,j}]=\Theta$

to display

.

Now, to type the same equation, my LaTeX source code looks like this:

$λ/μ=κ/ν ⇒ ∃Θ,∀i, ∑_{j∈ℕ} E[D_{i,j}]=Θ$

which produces exactly the same output. The source code is much easier to read; it is also slightly easier to type. Here is how the magic works:

- In the preamble, add
\usepackage[utf8x]{inputenc} \usepackage{amssymb}

- A number of special characters (including all Greek letters) were already easily available to me because I use a bépo keyboard (if you are a French speaker, you should try it out); otherwise, all characters are available using any keyboard to users of a Unix-like OS thanks to this great .XCompose file. For example, to get ℕ, use the keys Compose+|+N (pretty intuitive, and faster than typing \mathbb{N}). To get ∃, use Compose+E+E; to get ∈, use Compose+i+n, and so on.
- There are two issues with this solution: first, the unicode symbol α maps to \textalpha instead of \alpha; second, the blackboard letters map to \mathbbm instead of \mathbb. This can lead to errors, but I wrote this file which solves the issue by including in the preamble:
\input{greektex.tex}

This is useful for LaTeX, but also for all other places where you might want to type math: thanks to this .XCompose file, typing math in a blog post, tweet or e-mail becomes easy (for example, this is the last blog post where I will use WordPress’ $latex syntax). And if there ever is a LaTeX formula that you cannot access from your keyboard, you can use a website such as unicodeit.net which converts LaTeX source code to unicode.

I originally heard about this on Christopher Olah‘s blog.

]]>

I know several colleagues whose official position is that if the report is handed in late, the student fails the course. This is fine if the report is *very* late, but it seems a bit drastic for a small delay (e.g. a few minutes!). The problem of course is to define *very* and *small* in the previous sentence. And even if you decide that the course is failed for students who are more than (say) 24 hours late, it seems a bit unfair to have such a big difference between being 23 hours 59 minutes late and 24 hours 1 minute…

This time, I am using a continuous decreasing function as penalty: I told the students that

“Your report should be sent by e-mail by 2 April 2013 at noon. In case of a late submission, your grade will be multiplied by exp(-0.001 ⋅ t³ /86400³ ) where t is the number of seconds between the deadline and your submission.”

This basically means that they can be 2 days late for free. If a student has a legitimate reason for being late, they will usually manage to fit in that window, so they do not need to explain their lateness. Being 4 days late knocks off 6% of the grade; being a week late knocks off 30%. A student more than 8.8 days late loses 50% of their grade, meaning the course is necessarily failed, even if the report is perfect.

In this case, the last student to hand in his work was about 3.5 days late, meaning the mark will be multiplied by 0.953.

I quite like this system, although I think it was a bit too generous with this exact implementation. I cannot claim to have invented it; I read about it somewhere else, with a different penalty function, but cannot find where.

Edit: Thanks to Mahendra for the original post: here, on the Messy Matters blog.

]]>

Three young and talented speakers have accepted to take part in this session, and I am looking forward to hearing what they have to say:

Alexis Muir-Watt will talk about PMCMC advances for the doubly-intractable problem of estimating a partial order, with application to the social order between 12th century bishops.

Lawrence Murray will also discuss PMCMC, for applications in the environmental sciences including marine biogeochemistry, soil carbon modelling and hurricane tracking.

Simon Barthelmé will discuss using quasi-Kronecker matrices to speed up MCMC for functional ANOVA problems, with an application in neurosciences.

All these speakers have been dealing with challenging data sets and models. These applications have led to methodological advances, and the session will at the same time showcase the variety of applications and the way they help BayesComp methodology progress.

MCMSki 4 will be held in Chamonix, January 6-8 2014.

]]>

1*. Statistical methods in Historical Linguistics *(Thursday morning)

Recent advances in our understanding of language change, in statistical methodology, and in computational power, along with an increasing wealth of available data, have allowed significant progress in statistical modelling of language change, and quantitative methods are gaining traction in Historical Linguistics. Models have been developed for the change through time of vocabulary, morpho-syntactic and phonetic traits. I shall present a review of these models (from a statistician’s point of view), starting with Morris Swadesh’s failed attempts at glottochronology, then looking at some models developed in the last decade. In parallel, I shall provide brief insights into statistical tools such as Bayesian statistics and Markov Chain Monte Carlo, in order to show how to use these effectively for linguistic applications.

2. *A phylogenetic model of language diversification *(Friday afternoon)

Language diversification is a random process similar in many ways to biological evolution. We model the diversification of so-called “core” lexical data by a stochastic process on a phylogenetic tree. We initially focus on the Indo-European language family. The age of the most recent common ancestor of these languages is of particular interest and issues of dating ancient languages have been subject to controversy. We use Markov Chain Monte Carlo to estimate the tree topology, internal node ages and model parameters. Our model includes several aspects specific to language diversification, such as rate heterogeneity and the data registration process, and we show that lexical borrowing does not bias our estimates. We show the robustness of our model through extensive validation and analyse two independent data sets to estimates the age of Proto-Indo-European. We then analyse a data set of Semitic languages, and show an extension of our model to explore whether languages evolve in “punctuational bursts”. Finally, we revisit an analysis of several small data sets by Bergsland & Vogt (1962).

Joint work with Geoff Nicholls

Joint work with Geoff Nicholls

3. *Tutorial and practical: TraitLab, a package for phylogenies of linguistic and cultural traits*

In this tutorial, I shall present how to use the TraitLab package, which was initially developed specifically for the modelling of core vocabulary change through time, and guide interested attendants through an analysis of a simple data set. TraitLab is a software package for simulating, fitting and analysing tree-like binary data under a stochastic Dollo model of evolution. It handles “catastrophic” rate heterogeneity and missing data. The core of the package is a Markov chain Monte Carlo (MCMC) sampling algorithm that enables the user to sample from the Bayesian joint posterior distributions for tree topologies, clade and root ages, and the trait loss and catastrophe rates for a given data set. Data can be simulated according to the fitted Dollo model or according to a number of generalized models that allow for borrowing (horizontal transfer) of traits, heterogeneity in the trait loss rate and biases in the data collection process. Both the raw data and the output of MCMC runs can be inspected using a number of useful graphical and analytical tools provided within the package. TraitLab is freely available and runs within the Matlab computing environment.

Attendants who wish to use TraitLab during the practical should have a computer with Matlab installed.

TraitLab was developed jointly with Geoff Nicholls and David Welch.

]]>