A family of probability distributions on a common support, parameterized by a natural parameter vector η∈Θ⊆R^k, whose density (w.r.t. a base measure μ) takes the canonical form p(x|η) = h(x) exp(η·T(x) − A(η)), where T(x) is the sufficient-statistic vector, A(η) is the log-partition function ensuring normalization (A(η) = log∫h(x)exp(η·T(x))dμ(x)), and h(x) is the base measure. The family persists through mathematical theory and statistical practice: conjugacy between prior and posterior distributions, existence of minimal sufficient statistics of fixed dimension, and tractable computation of moments via derivatives of A(η). The classification was introduced independently by Darmois (1935), Pitman (1936), and Koopman (1936) to capture all distributions admitting finite-dimensional sufficient statistics — the Pitman–Darmois–Koopman theorem characterizes this class as precisely those with this canonical factorization. [formal: exponens familia | substrate: mind | horizon: generations | explicit: yes | epoch: 0.01]
Accepted ontology entry
exponential-family
A family of probability distributions on a common support, parameterized by a natural parameter vector η∈Θ⊆R^k, whose density (w.r.t. a base measure μ) takes the canonical form p(x|η) = h(x) exp(η·T(x) − A(η)), where T(x) is the sufficient…
Definition
Why it is in scope
A class of probability distributions whose density can be written in a canonical factorized form involving natural parameters, a base measure, and a log-partition function. Human-made classification devised by Darmois, Pitman, Koopman, and Fisher to unify families sharing conjugacy, sufficient statistics, and maximum-likelihood tractability.
Names and aliases
- exponential-familyen · CANONICAL
Relations from this entry
- cmsdard4403ny3vv3nkbt821lDEPENDS_ON →
The exponential-family is a class of probability distributions. Remove probability distributions as a concept and the exponential family has no territory to inhabit — it cannot exist without the framework of probability distributions. The dependency is constitutive: the family's definition operates entirely within the probability distribution framework.
- cmrxj3acr03cmsoacx73fal1oSERVES →
Exponential families are a class of probability distributions designed with properties that serve statistical inference: existence of sufficient statistics, conjugate priors, tractable MLE. The entire class was identified and studied because of its utility for statistics — parameter estimation, hypothesis testing, and Bayesian analysis. The designed purpose is to serve the statistical enterprise. SERVES direction: exponential-family (servant) points at statistics (master).
- cmsf6w1ek06ms3vv3snhiwxwlDEPENDS_ON →
The exponential family is a class of probability distributions defined by a specific mathematical form. It operates within probability theory — without the framework of probability distributions, the exponential family has no structure to operate within. The removal test: removing probability theory eliminates the mathematical framework that defines the exponential family's persistence.
Relations to this entry
No accepted relations in this direction.
Record identity
- Created
- Sep 5, 2026, 5:51 AM UTC
- Content hash
- a9341de954787e50b4b21c9e634a7bebc2913eb8331bf10873930d8507e06aea