This was the second viva of the week, for the thesis of Madeleine Thompson, but as it was in Toronto, I took part in it by a phone connection. This was rather ineffective as the connection was rather poor and I could not follow most of the questions… I had previously read (and commented) two papers, Slice Sampling with Adaptive Multivariate Steps: The Shrinking-Rank Method, and Graphical Comparison of MCMC Performance, co-written by Madeleine so I was well-aware of a part of the contents of the thesis, which I read in toto a few weeks ago. It was an interesting thesis with diversified threads in the various chapter, but I found frustrating to be unable to fully take part in the thesis debate… In retrospect, I should have flown to Toronto from Manchester yesterday or abstained from taking part in the viva!
Archive for PhD thesis
Both viva and talk went on well (even though I was a bit too tired to give a good talk, I fear!), with interesting outcomes in both cases. The viva lasted over two hours with an exciting exchange over the increase in overall error linked with the increase in dimension and over handling HMMs with four parameters to calibrate in parallel. At some point I got confused with Dennis’ result that
which I though was contradicting my favourite example of the non-central chi-square domination of the regular normal mean, namely that
under squared error loss. (This is Example 3.35 in The Bayesian Choice.) I had completely forgotten that the Jeffreys’ priors associated with both posterior expectations were different! The above equality is thus not invalidated by this example. It is further quite interesting in that it shows the posterior expectation is a sort of weak sufficient statistics for the estimation of the parameter, even though I remain in favour of using more summary statistics in ABC than a posterior expectation or a pseudo-MLE. In any case, the discussion of the corresponding Read Paper at the Royal Statistical Society next December 14 promises to be interesting and well-attended… Overall, the trip was quite pleasant (nice hotel, nice run in the countryside, where I took the attached pictures) and profitable, with discussions with Paul Fearnhead gearing me towards taking advantage of my colleagues’ expertise on indirect inference at CREST.
Yesterday, I flew to València to take part in Anabel Forte’s thesis defence, her thesis being on Objective Bayes criteria for variable selection and involving hyper-g priors. The view over the Pyrénées was quite spectacular, even though there was not much snow. Further South, I also liked this contrast between the red arid land and the river in the middle:
On Wednesday, I went to the University of Warwick to take part in a meeting about their new MASDOC programme. This programme was launched last year with the support of the EPSRC in three U.K. universities, Warwick, Lancaster and Cambridge (for math). It prepares graduate (fifth year) students for conducting a PhD in Mathematics or Statistics by providing them with extra tutoring and by creating a “cohort” of students working together on research topics. For the first year, the cohort was made of eleven students selected among applicants from both the UK and abroad. Besides a solid volume of courses in Mathematics, Probability and Statistics, MASDOC has the students working in small teams on an applied math problem (e.g., data assimilation, biomembranes, brain imaging) in order to (a) determine a reserarch programme and (b) propose a solution. The teams switch between (a) and (b) which is a neat good idea. The students are also given a common working room in order to increase their team abilities. When discussing with them, I was quite impressed by their maturity and involvement, as they already had a vision of their research interests. In fact, they have somehow gained one year ahead of the average student in terms of decision-making and planning, if not in terms of contents. Of course, this approach to graduate training is rather elitist in that it cannot be extended to all first-year graduates, however it is a worthy investment by EPSRC and the selected universities for building a core of PhD students and future academics with a broader spectrum, a more mature approach to research and teamwork, and hence a higher efficiency now and later. As a side issue, the MASDOC programme is also pushing for exchanges between institutions at the graduate and PhD levels, which is always a plus. Especially when considering the possibilities offered by the Paris graduate school of mathematical sciences.
I received this question from Luke Bornn to answer for a new Q&A entry in the ISBA Bulletin:
“From your experience, what skill do you think is most often lacking in today’s statistics Ph.D. graduates? What steps can a current graduate student undertake to remedy this deficiency?”
First, let me stress that I restrict my answer to French Ph.D. graduates and warn the reader that the environment for Ph.D. students in French institutions strongly differs from the ones in UK or US universities. Even though our students have a proper five-year training in maths, probability and statistics (plus possibly additional fields like economics, computer science, engineering, sociology, or, more rarely, biology, astronomy, ecology), there is not the same progressive integration of graduate students within the research faculty body as the one we see in the UK or the US. Ph.D. students remain students till the end of their thesis and often beyond. This is of course a terrible situation that we are trying to alleviate at our individual level, when the conditions allow as in CREST.