**A** chance question on X validated made me reconsider about the minimaxity over the weekend. Consider a Geometric G(p) variate X. What is the minimax estimator of p under squared error loss ? I thought it could be obtained via (Beta) conjugate priors, but following Dyubin (1978) the minimax estimator corresponds to a prior with point masses at ¼ and 1, resulting in a constant estimator equal to ¾ everywhere, except when X=0 where it is equal to 1. The actual question used a penalised qaudratic loss, dividing the squared error by p(1-p), which penalizes very strongly errors at p=0,1, and hence suggested an estimator equal to 1 when X=0 and to 0 otherwise. This proves to be the (unique) minimax estimator. With constant risk equal to 1. This reminded me of this fantastic 1984 paper by Georges Casella and Bill Strawderman on the estimation of the normal bounded mean, where the least favourable prior is supported by two atoms if the bound is small enough. Figure 1 in the Negative Binomial extension by Morozov and Syrova (2022) exploits the same principle. (Nothing Orwellian there!) If nothing else, a nice illustration for my Bayesian decision theory course!

## Archive for Geometric distribution

## a [counter]example of minimaxity

Posted in Books, Kids, Statistics, University life with tags 1984, cross validated, Geometric distribution, George Orwell, least favourable priors, minimaxity, negative binomial distribution, Statistical Decision Theory and Bayesian Analysis, The Bayesian Choice on December 14, 2022 by xi'an## Buffon machines

Posted in Books, pictures, Statistics, University life with tags Bernoulli factory, Buffon machines, Buffon's needle, Geometric distribution, George Louis Leclerc, John von Neumann, logarithmic distribution, Luc Devroye, Poisson distribution on December 22, 2020 by xi'anBy chance I came across a 2010 paper entitled On Buffon Machines and Numbers by Philippe Flajolet, Maryse Pelletier and Michèle Soria. Which relates to Bernoulli factories, a related riddle, and the recent paper by Luis Mendo I reviewed here. What the authors call a *Buffon machine* is a device that produces a perfect simulation of a discrete random variable out of a uniform bit generator. Just like (George Louis Leclerc, comte de) Buffon’s needle produces a Bernoulli outcome with success probability π/4. out of a real Uniform over (0,1). Turned into a sequence of Uniform random bits.

*“Machines that always halt can only produce Bernoulli distributions whose parameter is a dyadic rational.”*

When I first read this sentence it seemed to clash with the earlier riddle, until I realised the later started from a B(p) coin to produce a fair coin, while this paper starts with a fair coin.

The paper hence aims at a version of the Bernoulli factory problem (see Definition 2), although the term is only mentioned at the very end, with the added requirement of simplicity and conciseness translated as a finite expected number of draws and possibly an exponential tail.

It first recalls the (Forsythe–)von Neumann method of generating exponential (and other) variates out of a Uniform generator (see Section IV.2 in Devroye’s generation bible). Expanded into a general algorithm for generating discrete random variables whose pmf’s are related to permutation cardinals,

if the Bernoulli generator has probability λ. These include the Poisson and the logarithmic distributions and as a side product Bernoulli distributions with some logarithmic, exponential and trigonometric transforms of λ.

As a side remark, a Bernoulli generator with probability 1/π is derived from Ramanujan identity

as “a discrete analogue of Buffon’s original. In a neat connection with Mendo’s paper, the authors of this 2010 paper note that Euler’s constant does not appear to be achievable by a Buffon machine.

## pure birth process

Posted in Statistics with tags birth process, differential equation, Easter, egg double yolks, Geometric distribution, The Riddler, Yule on April 18, 2020 by xi'anThe Riddler has a rather simplistic riddle this week since it essentially asked for the expectation of a pure birth process (also known as the Yule process) at time t. Since the population size at time *t* has a geometric distribution with expectation

*e ^{λt}*.

It however took me a while to recover this result on my own on Easter afternoon, as I went for the integrals rather than the distribution itself and the associated differential equations. Interestingly (in a local sense!), I first following the wrong path of looking at the average time to the first birth, 1/*λ*, then to the second, 2/*λ*, and so on. Wrong since of course expectations do not carry this way… For a unit rate, *λ*=1, the average time to reach 10 births is about 3, while the average number of births over t=3 is essentially 20.

## survivalists [a Riddler’s riddle]

Posted in Books, Kids, R, Statistics with tags Geometric distribution, mathematical puzzle, order statistics, recurrence, riddle, survival analysis, The Riddler on April 22, 2019 by xi'an**A** neat question from The Riddler on a multi-probability survival rate:

Nine processes are running in a loop with fixed survivals rates .99,….,.91. What is the probability that the first process is the last one to die? Same question with probabilities .91,…,.99 and the probability that the last process is the last one to die.

The first question means that the realisation of a Geometric G(.99) has to be strictly larger than the largest of eight Geometric G(.98),…,G(.91). Given that the cdf of a Geometric G(a) is [when counting the number of attempts till failure, included, i.e. the Geometric with support the positive integers]

the probability that this happens has the nice (?!) representation

which leads to an easy resolution by recursion since

and

and a value of 0.5207 returned by R (Monte Carlo evaluation of 0.5207 based on 10⁷ replications). The second question is quite similar, with solution

and value 0.52596 (Monte Carlo evaluation of 0.52581 based on 10⁷ replications).

## standard distributions

Posted in Books, Kids, Statistics with tags arXiv, Bernoulli distribution, Geometric distribution, normal distribution on February 5, 2016 by xi'anJoram Soch managed to get a short note arXived about the Normal cdf Φ by exhibiting an analytical version, nothing less!!! By which he means a power series representation of that cdf. This is an analytical [if known] function in the complex calculus sense but I wonder at the point of the (re)derivation. (I do realise that something’s wrong on the Internet is not breaking news!)

Somewhat tangentially, this reminds me of a paper I read recently where the Geometric Geo(p) distribution was represented as the sum of two independent variates, namely a Binomial B(p/(1+p)) variate and a Geometric 2G(p²) variate. A formula that can be iterated for arbitrarily long, meaning that a Geometric variate is an infinite sum of [powers of two] weighted Bernoulli variates. I like this representation very much (although it may well have been know for quite a while). However I fail to see how to take advantage of it for simulation purposes. Unless the number of terms in the sum can be determined first. And even then it would be less efficient than simulating a single Geometric…