Questions tagged [exponential-family]

A set of distributions (eg, normal, $\chi^2$, Poisson, etc) that share a specific form. Many of the distributions in the exponential family are standard, workhorse distributions in statistics, w/ convenient statistical properties.

In probability and statistics, the exponential family is an important class of probability distributions sharing a certain form, specified below. This special form is chosen for mathematical convenience, on account of some useful algebraic properties, as well as for generality, as exponential families are in a sense very natural distributions to consider. The term exponential class is sometimes used in place of "exponential family".

From Mood et al. (pages 312 and 313, 1974):

  1. One-parameter exponential family.

A one-parameter family ($\theta$ is unidimensional) of densities $f(.;\theta)$ that can be expressed as:

$f_X(x;\theta) = \text{a}(\theta)\text{b}(x)\text{exp}[\text{c}(\theta)\text{d}(x)]$,

for $-\infty < x < \infty$, for all $\theta \in$ parametric space, and for a suitable choice of functions $\text{a}(.),\text{b}(.),\text{c}(.)$, and $\text{d}(.)$ is defined to belong to the exponential family or exponential class.

  1. K-parameter exponential family.

A family of densities $f(.,\theta_1,...,\theta_k)$ that can be expressed as:

$f_X(x;\theta_1,...,\theta_k) = \text{a}(\theta_1,...,\theta_k)\text{b}(x)\text{exp}\sum\limits_{j=1}^k{\text{c}_j(\theta_1,...,\theta_k)\text{d}_j(x)}$,

for a suitable choice of functions $\text{a}(.,...,.), \text{b}(.), \text{c}_j(.,...,.)$, $\text{d}_j(.)$, $j=1,....,k$, is defined to belong to the exponential family.

References:

Mood, A. M., Graybill, F. A., & Boes, D. C. (1974). Introduction to theory of statistics. (B. C. Harrinson & M. Eichberg, Eds.) (3rd ed., p. 564). McGraw-Hill, Inc.

Wkipedia

361 questions
32
votes
4 answers

Why is the exponential family so important in statistics?

Why is the exponential family so important in statistics? I was recently reading about the exponential family within statistics. As far as I understand, the exponential family refers to any probability distribution function that can be written in…
stats_noob
  • 5,882
  • 1
  • 21
  • 42
29
votes
1 answer

When if ever is a median statistic a sufficient statistic?

I came across a casual remark on The Chemical Statistician that a sample median could often be a choice for a sufficient statistic but, besides the obvious case of one or two observations where it equals the sample mean, I cannot think of another…
Xi'an
  • 90,397
  • 9
  • 157
  • 575
21
votes
2 answers

Advantages of the Exponential Family: why should we study it and use it?

So here I am studying inference. I would like that someone could enumerate the advantages of the exponential family. By exponential family, I mean the distributions which are given as \begin{align*} f(x|\theta) = h(x)\exp\left\{\eta(\theta)T(x) -…
user242554
21
votes
3 answers

Why doesn't the exponential family include all distributions?

I am reading the book: Bishop, Pattern Recognition and Machine Learning (2006) which defines the exponential family as distributions of the form (Eq. 2.194): $$ p(\mathbf x|\boldsymbol \eta) = h(\mathbf x) g(\boldsymbol \eta) \exp \{\boldsymbol…
becko
  • 3,298
  • 1
  • 19
  • 36
20
votes
0 answers

Is there a general expression for ancillary statistics in exponential families?

An i.i.d sample $X_1,\dots,X_n$ from a scale family with c.d.f. $F(\frac{x}{\sigma})$ has $S(X)$ as an ancillary statistic if $S(X)$ depends on the sample only through $\frac{X_1}{X_n},\cdots,\frac{X_{n-1}}{X_n}$. Is this result also sufficient? Is…
19
votes
2 answers

Does log likelihood in GLM have guaranteed convergence to global maxima?

My questions are: Are generalized linear models (GLMs) guaranteed to converge to a global maximum? If so, why? Furthermore, what constraints are there on the link function to insure convexity? My understanding of GLMs is that they maximize a…
16
votes
3 answers

Poisson is to exponential as Gamma-Poisson is to what?

A Poisson distribution can measure events per unit time, and the parameter is $\lambda$. The exponential distribution measures the time until next event, with the parameter $\frac{1}{\lambda}$. One can convert one distribution into the other,…
16
votes
2 answers

Kullback–Leibler divergence between two gamma distributions

Choosing to parameterize the gamma distribution $\Gamma(b,c)$ by the pdf $g(x;b,c) = \frac{1}{\Gamma(c)}\frac{x^{c-1}}{b^c}e^{-x/b}$ The Kullback-Leibler divergence between $\Gamma(b_q,c_q)$ and $\Gamma(b_p,c_p)$ is given by [1]…
15
votes
3 answers

Definition of family of a distribution?

Does a family of a distribution have a different definition for statistics than in other disciplines? In general, a family of curves is a set of curves, each of which is given by a function or parametrization in which one or more of the parameters…
Carl
  • 11,532
  • 7
  • 45
  • 102
15
votes
2 answers

Derivation of normalizing transform for GLMs

$\newcommand{\E}{\mathbb{E}}$How is the $A(\cdot) = \displaystyle\int\frac{du}{V^{1/3}(\mu)}$ normalizing transform for the exponential family derived? More specifically: I tried to follow the Taylor expansion sketch on page 3, slide 1 here but…
14
votes
4 answers

What is the rationale behind the exponential family of distributions?

From elementary probability course, the probability distributions such as Gaussian, Poisson or exponential all have a good motivation. After staring at the formula of the exponential family distributions for a long time, I still do not get any…
13
votes
1 answer

Aside from the exponential family, where else can conjugate priors come from?

Do all conjugate priors have to come from the exponential family? If not, what other families are known to have/produce conjugate priors?
Josh
  • 3,408
  • 4
  • 22
  • 46
13
votes
1 answer

ML estimate of exponential distribution (with censored data)

In Survival Analysis, you assume the survival time of a r.v. $X_i$ to be exponentially distributed. Considering now that I have $x_1,\dots,x_n$ "outcomes" of i.i.d r.v.'s $X_i$. Only some proportion of these outcomes are in fact "fully realized",…
12
votes
1 answer

Do the mean and the variance always exist for exponential family distributions?

Assume a scalar random variable $X$ belongs to a vector-parameter exponential family with p.d.f. $$ f_X(x|\boldsymbol \theta) = h(x) \exp\left(\sum_{i=1}^s \eta_i({\boldsymbol \theta}) T_i(x) - A({\boldsymbol \theta}) \right) $$ where ${\boldsymbol…
Wei
  • 345
  • 1
  • 8
12
votes
1 answer

Does a canonical link function always exist for a Generalized Linear Model (GLM)?

In GLM, assuming a scalar $Y$ and $\theta$ for the underlying distribution with p.d.f. $$f_Y(y | \theta, \tau) = h(y,\tau) \exp{\left(\frac{\theta y - A(\theta)}{d(\tau)} \right)}$$ It can be shown that $ \mu = \operatorname{E}(Y) = A'(\theta)$. If…
Wei
  • 345
  • 1
  • 8
1
2 3
24 25