## the god delusion [statistically speaking]

Posted in Books, Kids, pictures, Statistics, Travel, University life with tags , , , , , on August 22, 2014 by xi'an

While in Bangalore, I spotted Richard Dawkins’ The God delusion in the [fantastic if chaotic] campus bookstore and bought the Indian edition for a nominal amount.  I read most of it during my week in Boston. And finished by the lake in Maine. While I agree with most of the points made in Dawkins’ book about the irrationality of religions, and of their overall negative impact on human societies, I found the first part rather boring in that I see little appeal in dissecting so minutely the [infinitely many] incoherences of religious myths and beliefs, as this will likely miss the intended target [i.e., literal believers]. Similarly, the chapter on evolution versus intelligent design made valuable points, albeit I had already seen them before. Nothing wrong with repeating those, in particular that evolution has little to do with chance, but again unlikely to convince the [fundamentalist] masses. Overall, the book mostly focus on the Judeo-Christian-Muslim branch of religions, which may reflect on the author’s own culture and upbringing but also misses the recent attempts of Buddhism to incorporate science into their picture.

“A universe in which we are alone except for other slowly evolved intelligences is a very different universe from one with an original guiding agent whose intelligent design is responsible for its very existence.” (p.85)

What is most interesting in the book (for me) is when Dawkins tries to set the God hypothesis as a scientific hypothesis and to apply scientific methods to validate or invalidate this hypothesis. Even though there is no p-value or quantitative answer at the end. Despite the highly frequent use of “statistical” and “statistically improbable” in the corresponding chapter. What’s even more fascinating is Dawkins’ take at Bayesian arguments! Either because it is associated with a reverent or because it relies on subjective prior assessments, Bayesian statistics does not fit as a proper approach. Funny enough, Dawkins himself relies on subjective prior probabilities when discussing the likelihood of find a planet such as Earth. Now, into the details [with the Devil1] in a rather haphazard order or lack thereof: Continue reading

## Bangalore workshop [ಬೆಂಗಳೂರು ಕಾರ್ಯಾಗಾರ] and new book

Posted in Books, pictures, R, Statistics, Travel, University life with tags , , , , , , , , , , , , on August 13, 2014 by xi'an

On the last day of the IFCAM workshop in Bangalore, Marc Lavielle from INRIA presented a talk on mixed effects where he illustrated his original computer language Monolix. And mentioned that his CRC Press book on Mixed Effects Models for the Population Approach was out! (Appropriately listed as out on a 14th of July on amazon!) He actually demonstrated the abilities of Monolix live and on diabets data provided by an earlier speaker from Kolkata, which was a perfect way to start initiating a collaboration! Nice cover (which is all I saw from the book at this stage!) that maybe will induce candidates to write a review for CHANCE. Estimation of those mixed effect models relies on stochastic EM algorithms developed by Marc Lavielle and Éric Moulines in the 90’s, as well as MCMC methods.

## Bangalore workshop [ಬೆಂಗಳೂರು ಕಾರ್ಯಾಗಾರ]

Posted in pictures, Running, Statistics, Travel, University life, Wines with tags , , , , , , , on August 3, 2014 by xi'an

As I am now back home after a rather lengthy and somewhat eventful trip [getting too early to Bangalore airport with 3 hours to spend in the nice and very quiet lounge, followed by another 5 hour wait in the very nice but no so quiet Bombay airport lounge, no visit to the cockpit this time!, and then the usual sick passenger blocking all trains from Paris-Charles de Gaulle airport for one hour, reaching home to find my 97-year old neighbour fallen in her kitchen and calling for help!], I cannot but reflect on the difference between my two trips to India, from the chaos of Varanasi to the orderly peace of the campus of the Indian Institute of Science of Bangalore and even to some extent of the whole city of Bangalore, all proportions guarded. Even managing to get a [new] pair of [new] prescription glasses (or rather spectacles) within three days!

I thus found this trip much less stressful and much profitable, from enjoying the local food to discussing with Indian statisticians. The purpose of the IFCAM workshop was to bring both groups together for potential joint projects funded by IFCAM (at the travel level). While I found most talks were driven by specific applications, esp. in genomics, there are directions where we could indeed collaborate, from capture-recapture to astrostatistics. So it may be that I’ll be back in India in a near future!

## Bangalore snapshot [ಬೆಂಗಳೂರು ಚಿತ್ರ]

Posted in pictures, Travel with tags , , , , , on August 2, 2014 by xi'an

## marriage with data

Posted in Kids, pictures, Travel with tags , , , , on August 1, 2014 by xi'an

## Bangalore snapshot [ಬೆಂಗಳೂರು ಚಿತ್ರ]

Posted in pictures, Travel with tags , , , , , on August 1, 2014 by xi'an

## Bangalore workshop [ಬೆಂಗಳೂರು ಕಾರ್ಯಾಗಾರ]

Posted in pictures, R, Running, Statistics, Travel, University life, Wines with tags , , , , , , on July 31, 2014 by xi'an

Second day at the Indo-French Centre for Applied Mathematics and the workshop. Maybe not the most exciting day in terms of talks (as I missed the first two plenary sessions by (a) oversleeping and (b) running across the campus!). However I had a neat talk with another conference participant that led to [what I think are] interesting questions… (And a very good meal in a local restaurant as the guest house had not booked me for dinner!)

To wit: given a target like

$\lambda \exp(-\lambda) \prod_{i=1}^n \dfrac{1-\exp(-\lambda y_i)}{\lambda}\quad (*)$

the simulation of λ can be demarginalised into the simulation of

$\pi (\lambda,\mathbf{z})\propto \lambda \exp(-\lambda) \prod_{i=1}^n \exp(-\lambda z_i) \mathbb{I}(z_i\le y_i)$

where z is a latent (and artificial) variable. This means a Gibbs sampler simulating λ given z and z given λ can produce an outcome from the target (*). Interestingly, another completion is to consider that the zi‘s are U(0,yi) and to see the quantity

$\pi(\lambda,\mathbf{z}) \propto \lambda \exp(-\lambda) \prod_{i=1}^n \exp(-\lambda z_i) \mathbb{I}(z_i\le y_i)$

as an unbiased estimator of the target. What’s quite intriguing is that the quantity remains the same but with different motivations: (a) demarginalisation versus unbiasedness and (b) zi ∼ Exp(λ) versus zi ∼ U(0,yi). The stationary is the same, as shown by the graph below, the core distributions are [formally] the same, … but the reasoning deeply differs.

Obviously, since unbiased estimators of the likelihood can be justified by auxiliary variable arguments, this is not in fine a big surprise. Still, I had not thought of the analogy between demarginalisation and unbiased likelihood estimation previously. Continue reading