Archive for Bruno de Finetti

ISBA 2016 [#2]

Posted in Books, pictures, Running, Statistics, Travel, University life, Wines with tags , , , , , , , , , , on June 15, 2016 by xi'an

Today I attended Persi Diaconis’ de Finetti’s ISBA Lecture and not only because I was an invited discussant, by all means!!! Persi was discussing his views on Bayesian numerical analysis. As already expressed in his 1988 paper. Which now appears as a foundational precursor to probabilistic numerics. And which is why I had a very easy time in preparing my discussion as I mostly borrowed from my NIPS slides. With some degree of legitimacy since I was already a discussant there. Anyway,  here is the most novel slide in the discussion, built upon my realisation that the principle behind nested sampling is fairly generic for integral approximation, rather than being restricted to marginal likelihood approximation.

persidiscussionAmong many interesting things, Persi’s talk made me think anew about infinite variance importance sampling. And about the paper by Souraj Chatterjee and Persi that I discussed a few months ago. In that some regularisation of those “useless” importance estimates can stem from prior modelling. Not as an aside, let me add I am very grateful to the ISBA 2016 organisers and to the chair of the de Finetti lecture committee for their invitation to discuss this talk!

on de Finetti’s instrumentalist philosophy of probability

Posted in Books, Statistics, Travel, University life with tags , , , , , , , , on January 5, 2016 by xi'an

Pont Alexandre III, Paris, May 8, 2012. On our way to the old-fashioned science museum, Palais de la Découverte, we had to cross the bridge on foot as the nearest métro station was closed, due to N. Sarkozy taking part in a war memorial ceremony there...On Wednesday January 6, there is a conference in Paris [10:30, IHPST, 13, rue du Four, Paris 6] by Joseph Berkovitz (University of Toronto) on the philosophy of probability of Bruno de Finetti. Too bad this is during MCMSkv!

De Finetti is one of the founding fathers of the modern theory of subjective probability, where probabilities are coherent degrees of belief. De Finetti held that probabilities are inherently subjective and he argued that none of the objective interpretations of probability makes sense. While his theory has been influential in science and philosophy, it has encountered various objections. In particular, it has been argued that de Finetti’s concept of probability is too permissive, licensing degrees of belief that we would normally call imprudent. Further, de Finetti is commonly conceived as giving an operational, behaviorist definition of degrees of belief and accordingly of probability. Thus, the theory is said to inherit the difficulties embodied in operationalism and behaviorism. We argue that these and some other objections to de Finetti’s theory are unfounded as they overlook various central aspects of de Finetti’s philosophy of probability. We then propose a new interpretation of de Finetti’s theory that highlights these central aspects and explains how they are an integral part of de Finetti’s instrumentalist philosophy of probability. Building on this interpretation of de Finetti’s theory, we draw some lessons for the realist-instrumentalist controversy about the nature of science.

Conditional love [guest post]

Posted in Books, Kids, Statistics, University life with tags , , , , , , , , , , , , , , , , , , , , on August 4, 2015 by xi'an

[When Dan Simpson told me he was reading Terenin’s and Draper’s latest arXival in a nice Bath pub—and not a nice bath tub!—, I asked him for a blog entry and he agreed. Here is his piece, read at your own risk! If you remember to skip the part about Céline Dion, you should enjoy it very much!!!]

Probability has traditionally been described, as per Kolmogorov and his ardent follower Katy Perry, unconditionally. This is, of course, excellent for those of us who really like measure theory, as the maths is identical. Unfortunately mathematical convenience is not necessarily enough and a large part of the applied statistical community is working with Bayesian methods. These are unavoidably conditional and, as such, it is natural to ask if there is a fundamentally conditional basis for probability.

Bruno de Finetti—and later Richard Cox and Edwin Jaynes—considered conditional bases for Bayesian probability that are, unfortunately, incomplete. The critical problem is that they mainly consider finite state spaces and construct finitely additive systems of conditional probability. For a variety of reasons, neither of these restrictions hold much truck in the modern world of statistics.

In a recently arXiv’d paper, Alexander Terenin and David Draper devise a set of axioms that make the Cox-Jaynes system of conditional probability rigorous. Furthermore, they show that the complete set of Kolmogorov axioms (including countable additivity) can be derived as theorems from their axioms by conditioning on the entire sample space.

This is a deep and fundamental paper, which unfortunately means that I most probably do not grasp it’s complexities (especially as, for some reason, I keep reading it in pubs!). However I’m going to have a shot at having some thoughts on it, because I feel like it’s the sort of paper one should have thoughts on. Continue reading

2013 WSC, Hong Kong

Posted in Books, pictures, Running, Statistics, Travel, University life with tags , , , , , , , , , on August 28, 2013 by xi'an

HongKong1After an early but taxing morning run overlooking the city, and a recovery breakfast (!), I went from my flat to the nearby Hong Kong Convention Centre where the ISI (2013 WSC) meeting is taking place. I had a few chats with friends and publishers (!), then read a chapter of Rissanen’s book over an iced coffee before attending the Bernoulli session. This was a fairly unusual session with a mix of history of probability, philosophy of probability and statistics, and computational issues (my talk). Edith Sylla gave some arguments as to why Ars Conjectandi (that she translated) was the first probability book ever. Krzys Burdzy defended his perspective on why von Mises and de Finetti were wrong (in their foundational views of statistics). And I gave my talk on a mixture of Bernoulli factory, Russian roulette and ABC  (After my talk, Victor Perez Abreu told me that Jakob Bernoulli had presumably used simulation to evaluate the variance of the empirical mean in the Bernoulli case.) What I found most interesting in the historical talk was that Bernoulli had proven his result in the late 1680’s but he waited to complete his book on moral and commercial issues, waited too long since he died before. This reminded me of Hume using probabilistic arguments a few years later to disprove the existence of miracles. And of Price waiting for Bayes’ theorem to counter Hume. The talk by Krzys was a quick summary of the views exposed in his book, which unsurprisingly did not convince me that von Mises and de Finetti (a) had failed and (b) needed to use a new set of (six) axioms to define probability. I often reflected on the fact that when von Mises and de Finetti state(d) that probability does not exist, they applied the argument to a single event and this does not lead to a paradox in my opinion. Anyway, this talk of Krzys’ induced most of the comments from the floor, my own talk being in fine too technical to fit in this historical session. (And then there was still some time to get to a tea shop in Sheng Wan to buy some Pu Ehr, if not the HK$3000 variety…!)

IMG_1561

Bayes 250th versus Bayes 2.5.0.

Posted in Books, Statistics, Travel, University life with tags , , , , , , , , , , , , , , , , , on July 20, 2013 by xi'an

More than a year ago Michael Sørensen (2013 EMS Chair) and Fabrizzio Ruggeri (then ISBA President) kindly offered me to deliver the memorial lecture on Thomas Bayes at the 2013 European Meeting of Statisticians, which takes place in Budapest today and the following week. I gladly accepted, although with some worries at having to cover a much wider range of the field rather than my own research topic. And then set to work on the slides in the past week, borrowing from my most “historical” lectures on Jeffreys and Keynes, my reply to Spanos, as well as getting a little help from my nonparametric friends (yes, I do have nonparametric friends!). Here is the result, providing a partial (meaning both incomplete and biased) vision of the field.

Since my talk is on Thursday, and because the talk is sponsored by ISBA, hence representing its members, please feel free to comment and suggest changes or additions as I can still incorporate them into the slides… (Warning, I purposefully kept some slides out to preserve the most surprising entry for the talk on Thursday!)

reading classics (#3)

Posted in Statistics, University life with tags , , , , , , , , , , , , on November 15, 2012 by xi'an

Following in the reading classics series, my Master students in the Reading Classics Seminar course, listened today to Kaniav Kamary analysis of Denis Lindley’s and Adrian Smith’s 1972 linear Bayes paper Bayes Estimates for the Linear Model in JRSS Series B. Here are her (Beamer) slides

At a first (mathematical) level this is an easier paper in the list, because it relies on linear algebra and normal conditioning. Of course, this is not the reason why Bayes Estimates for the Linear Model is in the list and how it impacted the field. It is indeed one of the first expositions on hierarchical Bayes programming, with some bits of empirical Bayes shortcuts when computation got a wee in the way. (Remember, this is 1972, when shrinkage estimation and its empirical Bayes motivations is in full blast…and—despite Hstings’ 1970 Biometrika paper—MCMC is yet to be imagined, except maybe by Julian Besag!) So, at secondary and tertiary levels, it is again hard to discuss, esp. with Kaniav’s low fluency in English. For instance, a major concept in the paper is exchangeability, not such a surprise given Adrian Smith’s translation of de Finetti into English. But this is a hard concept if only looking at the algebra within the paper, as a motivation for exchangeability and partial exchangeability (and hierarchical models) comes from applied fields like animal breeding (as in Sørensen and Gianola’s book). Otherwise, piling normal priors on top of normal priors is lost on the students. An objection from a 2012 reader is also that the assumption of exchangeability on the parameters of a regression model does not really make sense when the regressors are not normalised (this is linked to yesterday’s nefarious post!): I much prefer the presentation we make of the linear model in Chapter 3 of our Bayesian Core. Based on Arnold Zellner‘s g-prior. An interesting question from one student was whether or not this paper still had any relevance, other than historical. I was a bit at a loss on how to answer as, again, at a first level, the algebra was somehow natural and, at a statistical level, less informative priors could be used. However, the idea of grouping parameters together in partial exchangeability clusters remained quite appealing and bound to provide gains in precision….

Bayes-250, Edinburgh

Posted in Mountains, Statistics, Travel, University life with tags , , , , , , , on September 6, 2011 by xi'an

       

On September 5-7, I am attending Bayes-250 at the University of Edinburgh. I arrived under a blazing sun that showed Arthur’s Seat in full glory! (I realised I had never seen it in the sun before!)

My talk is a short version of the ABC in London talk about ABC model choice, in connection with the recently published PNAS paper.

On Monday afternoon, David Dunson gave a plenary talk on the use of tensors in Bayesian non-parametric modeling and the possibilities offered by this highly malleable tool that made me look forward reading his 2009 JASA paper with Xing. In the evening, David Spiegelhalter delivered a great (and well-attended) public lecture on the Bayesian nature of risk that made me think further about the de Finetti‘s “probability does not exist”, a declaration that had irked me until then. (And on which I will expand in a later comment on Error and Inference.) We ended up the day in Blonde, a nice restaurant we had already enjoyed during the mixture meeting in 2009.