Archive for Les Courants de la Liberté

just 39km/h faster…

Posted in pictures, Running, Travel with tags , , , , , , , , , , , , , on July 10, 2025 by xi'an

Pegasus bridge half marathon [1:26:26, 114/5227, 1/109 M5M and first over 55]

Posted in pictures, Running, Travel with tags , , , , , , , , , , , , , , on June 22, 2025 by xi'an


Tough race when compared with last year despite good weather conditions (if quite warm), and bagpipes at the start, maybe due to overtraining. I barely managed to keep my first position in the M5M category, by a mere 2s, not that I was aware of the threat to my crown! The briefly leased Hoka One One shoes did not particularly help this time (or maybe they did!). This was my last half marathon in the M5M category (if my first race on invitation) and I am thus glad I could end up on a winning note, if not on a podium since only scratch winners are rewarded that way. Hopefully, I will get re-invited for next year.

BayesComp 2025.1

Posted in Running, Statistics, Travel, University life with tags , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , on June 18, 2025 by xi'an

Minus one day at BayesComp 2025! As I am attending the model misspecification satellite workshop (ten minutes late, due to repeated path finding protocol!), with an extended presentation by Jeremias Knoblauch on post-Bayesian inference, incl. powered likelihood and Gibbs posteriors. A very smooth and pedagogical presentation, esp. in the hybrid mode. A perspective I associate with the difficulties of making sense of the post-posterior, not truly a posterior, of calibrating the penalty (eg λ), picking the loss (α, β, γ divergences?) , and the drift towards learning goals since the new measure is the post-posterior predictive. Sort of paradoxical return to a Gaussian post-posterior on the parameter that does not seem to stay robust. Horrendous computational issues, when the loss itself is an integral. Use of the zig-zag sampler with an estimated unbiased gradient of the loss, much faster than pseudo-marginal, which (naïvely?) makes sense both because PDMPs directly use scores and because of the power of stochastic gradient methods. Worse perspectives for optimisation-centric posterior that are essentially vamped versions of GANs. For instance, what is the meaning of the coverage probabilities?

The second talk by Jonathan Huggins was on DC (not bagged) posteriors as martingale posteriors with m<∞ (approximating marginal distributions with random kernel MCMC—which persists in simulating the marginalised or integrated variable u from its prior, rather than adapting to the current value of the parameter θ— or subsampling MCMC akin to stochastic gradient Langevin) with connection with cut posteriors,

Then I skipped to the second workshop on Bayesian methods for distributional and semiparametric regression, to listen to my friend David Rossell’s talk on local variable selection. Which suffers more than in standard models under misspecification. Another talk involving cut posteriors, the cuts being on the spline bases…

The day and the workshop concluded with great talks by (my friends) Pierre Alquier and David Frazier. David centred his misspecification talk on cut posteriors. Managing to bring in shrinkage estimators (and mention Bill Strawderman!).

A wee stressful trip, since the races in Caen cancelled all buses and delayed the taxi enough to miss the train to Paris by 30s, catching the next available one leaving me less than one hour between the arrival of the train (delayed by construction work on the rail line) and boarding the flight at Charles de Gaulle airport, but fortunately the RER trains in Paris were running okay, there were no queues in the airport, and I thus made it in time with a bit of post-marathon jogging! (Only to be delayed at departure by one hour for stormy conditions over Germany and Austria). All this exercise proved helpful to sleep soundly and lengthily in the plane!

2024 Huckleberry Running Festival 10K [39:28, 2nd/97, +174m, 18⁰]

Posted in Mountains, pictures, Running, Travel with tags , , , , , , , , , , , , , , , , , on August 16, 2024 by xi'an

A chance glance at a local newspaper, while checking for path conditions in the Zigzag Ranger station, made me aware of a local race taking place next to our vacation rental in Mount Hood Village! I first registered for the 5K then upgraded for the 10K, as the half-marathon sounded too much, esp. after an 18K fast training three days earlier… The route was nice, with two big hills and trees all around. I kept the first runner in my radar for the first seven kilometres but he then took off and won by almost two minutes. Very happy with the result, though, despite running the 5K and 10K more slowly than in Caen last month, where the altitude differential was much more modest. Too bad the tee-shirt is not that great! But the organising team was very friendly and excited at having an international participant to the races, for the very first time.

D Day Pegasus bridge half-marathon [1:23:42, 90th / 4883, 1st over 55, 13⁰, 117+m]

Posted in Kids, pictures, Running, Travel with tags , , , , , , , , , , on June 4, 2024 by xi'an

I had a great time at the Pegasus half-marathon, except for the freezing hundred minutes stuck at the starting point on one side of the bridge before the race started, when I had to escape the cold and the wind by staying between two edges! Great start over this historical (if since rebuilt) bridge (much shorter than the Golden Gate bridge!) accompanied by bagpipes, if not from a large band!, not too many runners to pass over the first kilometre thanks to the division into several waves, even though I was not strategically located close enough to the starting line, and quickly falling into a sub-4 pace that I kept for most of the race. Great that my family was there at the arrival line. Despite the proximity to the Channel and the lack of protection through the half-grown wheat fields, the wind was not too much of an hindrance and the 110m overall climb was dispersed enough over most of the race hence never impacting my pace for long. A major plus, the final 5k’s were mostly downhill which helped a lot with this usually challenging part (even though I do not look particularly serene on the pics)! I essentially stayed with the same runners over the second half of the race, only for a few young ‘uns passing me in the last 300m within the stadium (which sits next to my high school, Lycée Malherbe). I was thus more than happy with my time (and with my higher volume training in the prior month, with a week at 100km), a minute faster than planned and two minutes earlier than in Boulogne, but did not expect any podium position in such a large and competitive race and only checked my ranking after getting home (my mother’s home) and showering, sort of “too late” for attending the award ceremony but I do not think I would have gone anyway. Next half planned in Argentan, end of September, where else?!