Sousaphonic graph!

Posted in Books, pictures, Statistics with tags , , , , , , , , , , on January 17, 2022 by xi'an

Guiana impressions [#1]

Posted in Books, Kids, Mountains, pictures, Running, Travel with tags , , , , , , , , , , , , , , , , , , , , , , , , on January 16, 2022 by xi'an

As our daughter Rachel has started her (five year) medical residency with a semester round in a French Guiana hospital, we took the opportunity of the Xmas break and of acceptable travel restrictions to visit her and the largest (and sole American) French departement for a week! This was a most unexpected trip that we enjoyed considerably.

While hot and humid is not my favourite type of weather (!) the weather remained quite tolerable that week, esp. when considering this was the start of the rain season (guiana means land of plentiful water in Arawak!) This made hiking on the (well-traced) paths in the local equatorial rain forest rather interesting, as the red soil is definitely muddy or worse. I however faced much less insects than I feared and mosquito bites were rare beyond the dawn and dusk periods. Plenty of birds, albeit mostly invisible. Except for the fantastic marshes of Kaw, where the variety of birds is amazing, including aras and toucans. Very muddy trails, did I mention it, but beautiful explosion of trees. Green everywhere.My first sight of a sloth was quite the treat, but I regret not spotting anteaters. Or a tapir. Swimming in the marshes of Kaw was great as well, with no worry from local caimans! Which we went spotting after nightfall. The place reminded me in several ways of Tonlé Sap lake, near Angkor.

Ate there an atipa bosco fish from the same place. Which has samurai armor. And two front legs to move outside water! As we had no say in what was served, we also ate paca meat in this restaurant, the agouti paca being a local rodent. Unfortunately because bush meat should not be served to tourists for fear of reducing the animal populations.

Visited several remains of former penal colonies, the whole country being a French penal colony at a not-so-distant-time, from the era when Louisiana was sold to the U.S. to the abolition in 1938, only implemented in 1953… Appalling to think that political and criminal prisoners were sent there to slowly rot to death, with no economical purpose on top of it! To the point of dead prisoners being immersed at sea rather than buried on island gallows, the local cemetery being reserved to guardians and their families….

Sandremonde [book review]

Posted in Statistics with tags , , , , , , on January 15, 2022 by xi'an

A somewhat original fantasy book (in French), by Jean-Luc Deparis, found by chance on a shelf of La Case à Bulles bookstore in Cayenne, with a nice if traditional cover, a first half I enormously enjoyed and read within a day!. and a second one less appealing to my tastes and which took me longer to complete, despite eventually skipping some passages… Presumably because this second part involved more magic, [dreaded!] endless subterranean domains, and the unsurprising revelation of a predestination for the central heroine, which was till then doing well by herself, thank you very much. Although with early and heavy hints of a unique destiny. The beginning has flavours reminding me of The Lies of Locke LamoraHobbs’ Assassin series, and the more recent Red/Gray/Holy Sister trilogy. Despite its flaws, the non-magical universe of Sandremonde is fascinating, with an overwhelming Church of monk-soldiers that has the monopoly of (magical) protections and as such promotes or demotes local lordlings… The end is both predictable and of little interest. More editing and advice from the publisher would have considerably improved the outcome. May a prospective second novel by the author keep the imagination and avoids the clichés!

health [s]care

Posted in Kids, pictures with tags , , , , , , , , , , , on January 14, 2022 by xi'an

prior elicitation

Posted in Books, Kids, Statistics, University life with tags , , , , , , , , , , , , , on January 13, 2022 by xi'an

“We believe that an elicitation method should support elicitation both in the parameter and observable space, should be model-agnostic, and should be sample-efficient since human effort is costly.”

Petrus Mikkola et al. arXived a long paper on prior elicitation addressing the (most relevant) question: Why are we not widely use prior elicitation? With a massive bibliography that could be (partly) commented (and corrected as some references are incomplete, as eg my book chapter on priors!). I think the paper would make a terrific discussion paper.

The absence of a general procedure for prior elicitation is indeed hindering the adoption of Bayesian methods outside our core community and is thus eventually detrimental to their wider development. It also carries the dangers of misled or misleading prior choices. The authors put forward the absence of “software that integrates well with the current probabilistic programming tools used for other parts of the modelling workflow.” This requires setting principles that avoid “just-press-key” solutions. (Aside: This reminds me of my very first prospective PhD student, who was then working in a startup [although the name was not yet in use in the early 1990’s!] and had build such a software in a discretised, low dimension, conjugate prior, environment by returning a form of decision-theoretic impact of the chosen hyperparameters. He alas aborted his PhD attempt due to the short-term pressing matters in the under-staffed company…)

“We inspect prior elicitation from the perspectives of (1) properties of the prior distribution itself, (2) the model family and the prior elicitation method’s dependence on it, (3) the underlying elicitation space, (4) how the method interprets the information provided by the expert, (5) computation, (6) the form and quantity of interaction with the expert(s), and (7) the assumed capability of the expert (…)”

Prior elicitation is indeed a delicate balance between incorporating expert opinion(s) and avoiding over-standardisation. In my limited experience, experts tend to be over-confident about their own opinion and unwilling to attach uncertainty to their assessments. Even when being inconsistent. When several experts are involved (as, very briefly, in Section 3.6), building a common prior quickly becomes a challenge, esp. if their interests (or utility functions) diverge. As illustrated in the case of the whaling commission analysed by Adrian Raftery in the late 1990’s. (The above quote involves a single expert.) Actually, I dislike the term expert altogether, as it comes without any grading of the reliability of the person.To hit (!) at an early statement in the paper (p.5), should the prior elicitation always depend on the (sampling) model, as experts may ignore or misapprehend the model? The posterior already accounts for the likelihood and the parameter may pre-exist wrt the model, as eg cosmological constants or vaccine efficiency… In a sense, the model should be involved as little as possible in the elicitation as the expert could confuse her beliefs about the parameter with those about the accuracy of the model. (I realise this is not necessarily a mainstream position as illustrated by this paper by Andrew and friends!)

And isn’t the first stumbling block the inability of most to represent one’s prior knowledge in probabilistic terms? Innumeracy is a shared shortcoming in the general population (and since everyone’s an expert!), as repeatedly demonstrated since the start of the Covid-19 pandemic. (See also the above point about inconsistency. Accounting for such inconsistencies in a Bayesian way is a natural answer, albeit requiring the degree of expertise and reliability to be tested.)

Is prior elicitation feasible beyond a few dimensions? Even when using the constrictive tool of copulas one hits a wall after a few dimensions, assuming the expert is willing to set a prior correlation matrix.  Most of the methods described in Section 3.1 only apply to textbook examples. In their third dimension (!), the authors mention neural network parameters but later fail to cover this type of issue. (This was the example I had in mind indeed.) And they move from parameter space to observable space. Distinguishing predictive elicitation from observational elicitation, the former being what I would have suggested from scratch. Obviously, the curse of dimensionality strikes again unless one considers summary statistics (like in ABC).

While I am glad conjugate priors do not get the lion’s share, using as in Section 3.3.. non-parametric or machine learning solutions to construct the prior sounds unrealistic. (And including maximum entropy priors into that category seems wrong since they are definitely parametric.)

The proposed Bayesian treatment of the expert’s “data” (Section 4.1) is rational but requires an additional model construct to link the expert’s data with the parameter to reach a Bayes formula like (4.1). Plus a primary prior (which could then be one of the reference priors.) Reducing the expert’s input to imaginary observations may prove too narrow, though. The notion of an iterative elicitation is most appealing and its sequential aspect may not be particularly problematic in opposition to posteriors relying on using the data twice or more. I am much less buying the hierarchical construct of Section 4.3 because they imply a return to conjugate priors and hyperpriors, are not necessarily correctly understood by experts, do not always cater to observational elicitation, and are not an answer to high-dimension challenges.

Given the state of the art, it sounds like we are still far from seeing prior elicitation as a natural part of Bayesian software and probabilistic programming. Even when using a modular, model-agnostic strategy. But this is most certainly a worthy prospect!

%d bloggers like this: