This morning, on my way to the airport (and to Montpellier for a seminar), Rock, my favourite taxi-driver, told me of a strange ride he endured the night before, so strange that he had not yet fully got over it! As it happened, he had picked an elderly lady with two large bags in the vicinity after a radio-call and drove her to a sort of catholic hostel in down-town Paris, near La Santé jail, a pastoral place housing visiting nuns and priests. However, when they arrived there, she asked the taxi to wait before leaving, quite appropriately as she had apparently failed to book the place. She then asked my friend to take her to another specific address, an hotel located nearby at Denfert-Rochereau. While Rock was waiting and the taxi counter running, the passenger literally checked in by visiting the hotel room and deciding she did not like it so she gave my taxi yet another hotel address near Saint-Honoré where she repeated the same process, namely visited the hotel room with the same outcome that she did not like the place. My friend was then getting worried about the meaning of this processionary trip all over Paris, the more because the lady did not have a particularly coherent discourse. And could not stop talking. The passenger then made him stop for food and drink, and, while getting back in the taxi, ordered him to drive her back to her starting place. After two hours and half, they thus came back to the place, with a total bill of 113 euros. The lady then handled a 100 euro bill to the taxi-driver, declaring she did not have any further money and that he should have brought her home directly from the first place they had stopped… In my friend’s experience, this was the weirdest passenger he ever carried and he thought the true point of the ride was to escape solitude and loneliness for one evening, even if chatting about non-sense the whole time.
Archive for Paris
My trip to work was somewhat more eventful than usual this morning: as the queue to switch to the A train was too long for my taste, I exited the Chatelet station to grab a Vélib rental bike near Le Louvre and followed the Louvre palace for a few hundred meters, until reaching a police barricade that left the remainder of the Rivoli street empty, a surreal sight on a weekday! As it happened, Beji Caid Essebsi, the president of Tunisia was on a state visit to Paris and staying at the 5-star Hotel Meurice. And just about to leave the hotel. So I hanged out there for a few minutes and watched a caravan of official dark cars leave the place, preceded by police bikes in formal dress! The ride to Dauphine was not yet straightforward as the Champs-Elysées had been closed as well, since the president was attending a commemoration (for Tunisian soldiers who died in French wars?) at Arc de Triomphe. This created a mess for traffic in the surrounding streets. Especially with pedestrians escaping from stuck buses and crowding my sidewalks! And yet another surreal sight of the Place de l’Étoile with no car. (In this end, this initiative of mine took an extra 1/2 hour on my average transit time…)
I am off to New York City for two days, giving a seminar at Columbia tomorrow and visiting Andrew Gelman there. My talk will be about testing as mixture estimation, with slides similar to the Nice ones below if slightly upgraded and augmented during the flight to JFK. Looking at the past seminar speakers, I noticed we were three speakers from Paris in the last fortnight, with Ismael Castillo and Paul Doukhan (in the Applied Probability seminar) preceding me. Is there a significant bias there?!
There is an opening at the Statistics School ENSAE for a Statistics associate or full professor position, starting on September 2015. Currently located on the South-West boundary of Paris, the school is soon to move to the mega-campus of Paris Saclay, near École Polytechnique, along with a dozen other schools. See this description of the position. The deadline is very close, March 23!
This morning, in the train to Dauphine (train that was even more delayed than usual!), I read a recent arXival of Brendon Brewer and Courtney Donovan. Entitled Fast Bayesian inference for exoplanet discovery in radial velocity data, the paper suggests to associate Matthew Stephens’ (2000) birth-and-death MCMC approach with nested sampling to infer about the number N of exoplanets in an exoplanetary system. The paper is somewhat sparse in its description of the suggested approach, but states that the birth-date moves involves adding a planet with parameters simulated from the prior and removing a planet at random, both being accepted under a likelihood constraint associated with nested sampling. I actually wonder if this actually is the birth-date version of Peter Green’s (1995) RJMCMC rather than the continuous time birth-and-death process version of Matthew…
“The traditional approach to inferring N also contradicts fundamental ideas in Bayesian computation. Imagine we are trying to compute the posterior distribution for a parameter a in the presence of a nuisance parameter b. This is usually solved by exploring the joint posterior for a and b, and then only looking at the generated values of a. Nobody would suggest the wasteful alternative of using a discrete grid of possible a values and doing an entire Nested Sampling run for each, to get the marginal likelihood as a function of a.”
This criticism is receivable when there is a huge number of possible values of N, even though I see no fundamental contradiction with my ideas about Bayesian computation. However, it is more debatable when there are a few possible values for N, given that the exploration of the augmented space by a RJMCMC algorithm is often very inefficient, in particular when the proposed parameters are generated from the prior. The more when nested sampling is involved and simulations are run under the likelihood constraint! In the astronomy examples given in the paper, N never exceeds 15… Furthermore, by merging all N’s together, it is unclear how the evidences associated with the various values of N can be computed. At least, those are not reported in the paper.
The paper also omits to provide the likelihood function so I do not completely understand where “label switching” occurs therein. My first impression is that this is not a mixture model. However if the observed signal (from an exoplanetary system) is the sum of N signals corresponding to N planets, this makes more sense.