Archive for ISBA 2026
Asahi-dake [jatp]
Posted in Mountains, pictures, Running, Travel with tags Asahi-dake, Biei, hiking, Hokkaido, ISBA 2026, Japan, Kamuy mosir, summitting, vacations, volcano on September 14, 2026 by xi'an
back from Japan [flotsam, jetsam, lagan, and derelict]
Posted in Books, Mountains, pictures, Running, Travel with tags Ainu, Air France, airport food, black bear, deep vein thrombosis, EU compensation rules, Haida Gwaii, Haneda, hiking, Hokkaido, hospital, ISBA 2026, Japan, Japan Airlines, Japanese Alps, jetlag, Krasheninnikov, Mont Bell, Mont-Saint-Michel, Nagano, Nagoya, North, outdoor, ronin, samurai, Sapporo, Tachikawa, TNF, Tokyo, volcano on August 8, 2026 by xi'an
Our last day in Japan was wrecked by Japan Airlines, as our JAL flight from Sapporo to Tokyo Haneda got delayed by five hours… The worst part was being kept in the dark about the delay till the initial departure time and then more as the actual departure time was only given hours later. An opportunity to find out that EU rules about flight compensations and passenger rights do not have a Japanese equivalent..! The only plus side of spending the day at the airport was enjoying a sashimi meal at the Sapporo airport. The next day, the erupting Krasheninnikov volcano in Kamtchatka caused a slight change of flight on the way (North!) back to Paris but Air France still managed a 45mn gain on the planned arrival time. And hardly any jetlag in the following days, as opposed to the unusually long accustomisation through the nights in Nagoya and Tachikawa.
This was the only major hitch of our trip, besides a rental car delay for missing the official translation of our driving licences. Visiting both the Japanese Alps and the island of Hokkaido was a fantastic opportunity, thanks to ISBA 2026. Along enjoying the views and hikes in both locations, I could not refrain from making a connection between the geographies and histories of the Ainus on Hokkaido and of the Haidas in the Pacific Northwest, as they both got colonised and forcibly integrated at about the same epoch, with salmon canning companies enslaving them. The settlers were demoted Union soldiers in one case and former samurais (tondenhei) in the other. Small towns on Hokkaido had some similarities with those of the American and Canadian West, with one-street layout and plenty of space between houses, except for these spaces being intensely cultivated as all of the valleys I saw during that trip. (But very little wildlife, with a single sorry-looking fox, two fleeing monkeys, a single, inquisitive, squirrel, a few deers and no bear.)
On the mountains (and volcanoes) trails, I was impressed by the involvement and preparedness of most Japanese hikers, often quite senior, in climbing volcanoes and facing bear encounters, which also made me discover a very nice local outdoor brand, Mont Bell. (Also finding out that TNF Japan© is a separate franchise with its own products.) And adopt the hiking uniform of wearing shorts and tights, which proved surprisingly comfortable!
Due to a thrombosis scare, I further had the opportunity to visit (again!) the local hospital in Omachi, Nagoya Prefecture, if getting there by bike rather than by water-ambulance. With hugely efficient processing despite the language barrier (except for the doctor with whom I briefly spoke of Mont-Saint-Michel!). And the conclusion of no thrombosis, phew!
la vie (bayésienne), mode d’emploi
Posted in Books, pictures, Travel, University life with tags Bayesian foundations, conference, foundation lectures, France, free jazz, ISBA, ISBA 2026, Japan, Nagoya, slides, Sylvia Richardson, WINC AICHI on August 1, 2026 by xi'an
ISBA⁶
Posted in pictures, Running, Statistics, Travel, University life with tags adaptive MCMC, curry, doubly intractable posterior, embarrassingly parallel MCMC, fish market, food poisoning, GPU, hitsumabushi, ISBA 2026, ISBA 2028, Japan, label switching, logo, MCMC, mirror workshop, mixtures of distributions, Nagoya Castle, noise contrasting estimation, prequential loss, quadrature, Shinkansen, simulated annealing, SMC, stereographic MCMC, udon, Valencia conferences, Venice, Wisconsin on July 5, 2026 by xi'an
After a final early morning run in the rising sun, reaching a point north of the city where I could see a nearby volcano (not Fuji-san!), I attended both morning on MCMC, with (again) a range of interesting, mostly novel, questions and solutions. With no overlap with his talk at mostly Monte Carlo last month, Sam Livingstone gave convincing motivations for using new tools for designing proper scaling (in the limit) for some adaptive MCMC. Charles Margossian’s talk was particularly exciting for pushing for single step MCMC when massively run in parallel, provided warm-up is over (enough). And Saif Syed discussing a new version of annealed SMC, using the variability of the estimated normalising constant as an assessment of the (although I could not catch how the tridimensional calibration was handled). I was less convinced by Kyle Kuang’s approach to overcome identifiability issues such as label switching, as it sounded too simple to be universally applicable. Bingjing Tang returned to the challenge of doubly intractable posteriors. With motivations from functional inference and a solution reminding me of noise contrastive estimation. until I spoke with the authors and realised it was much closer to our recent paper with Edoardo and Julien. While Bjorn Sprungk’s talk on Metropolized interacting particle sampling reminded me of our pinball sampler, presented… 30 years ago at the 1996 Valencia meeting! But using the product of posteriors as a target sounds suboptimal when a target that would keep particles apart (with the correct marginals) would prove more exploratory.
The noon break was the last opportunity to sample one of the food stalls in the fish market nearby, with a spicy curry udon bowl. Cutting the eel addiction!
My final session—before catching a shinkansen to Tokyo for the Information Geometry, Privacy and Monte Carlo ISBA Satellite Meeting at the Institute of Statistical Mathematics—was about loss-based posteriors, with our PhD student Shreya Roy presenting her work on prequential posteriors. And Kshitij Khare on using a loss that allows for a regular Gibbs sampler implementation via a pseudo-model and consistency properties.
This cuvée of ISBA World Meeting was exceptionally (gouleyante and) enjoyable (except for my recurrent sleeping issues) from the diverse and well-balanced programme, to the choice of plenary speakers, to the practicality of the conference centre (except for the queues for the lift!) and its location in Nagoya, with its own, unsuspected, perks! With no food poisoning this time!! ISBA 2028 is scheduled to take place in Milwaukee and I am very unlikely to attend, unless a rogue mirror pops up!
ISBA⁵
Posted in pictures, Running, Statistics, Travel, University life with tags Aussois, Bayes at CIRM, BayesComp 2027, Bayesian predictive, Chamonix, committee on the future of ISBA conferences, cut models, ISBA 2021, ISBA 2026, Japan, Laplace approximation, Marseille, MCMSki IV, mirror workshop, misspecified model, multi-hub conference, Nagoya Castle, skew-symmetric distribution, Susie Bayarri, typhoon on July 4, 2026 by xi'an
After a pleasant run under the pouring rain (noisy rain that had not helped with my sleep or lack thereof), I attended both morning episodes of Calibrated Bayes, with a range of interesting, mostly novel, questions and solutions. And rekindling my interrogations about overparameterised models and model misspecification within a Bayesian framework. And the elusive notion of outliers.
The noon discussion around our report as the committee on the future of ISBA conferences went on rather well, given the circumstances, with about 50 participants with ideas and opinions on the opportunity of splitting the ISBA World into multiple hubs. And on the practical difficulties. (Despite the itch to do it, I did not intervene with my counter-objections to most of the raised objections! Nor mentioned BayesComp in Aussois.) Participants of the (unofficial) 2021 mirror in Marseille brought some welcomed support!
Daniele Durante’s Bayarri lecture on skew symmetric approximations was also attuned to the approximation spirit of the morning. The symmetrisation of the target reminded me of our (unpublished) folded method. With the (unrelated) question of the statistical meaning of a higher order Laplace expansion. And the calibration of this skewed approximation.
The final session on Cut Bayes couldn’t be missed!, since this has been an interest of mine’s since MCMSki IV in Chamonix! With an automated cut model construction by Robert Goude and an optimised choice of generalised Bayes posteriors by David Nott.
