web.archive.org

Dirichlet distribution: Information from Answers.com

Several images of the probability density of the Dirichlet distribution when K=3 for various parameter vectors α. Clockwise from top left: α=(6, 2, 2), (3, 7, 5), (6, 2, 6), (2, 3, 4).

In probability and statistics, the Dirichlet distribution (after Johann Peter Gustav Lejeune Dirichlet), often denoted Dir(α), is a family of continuous multivariate probability distributions parametrized by the vector α of positive reals. It is the multivariate generalization of the beta distribution, and conjugate prior of the multinomial distribution in Bayesian statistics. That is, its probability density function returns the belief that the probabilities of K rival events are x_i given that each event has been observed α_i − 1 times.

Probability density function

Illustrating how the log of the density function changes when K=3 as we change the vector α from α=(0.3, 0.3, 0.3) to (2.0, 2.0, 2.0), keeping all the individual α_i's equal to each other.

The Dirichlet distribution of order K ≥ 2 with parameters α₁, ..., α_K > 0 has a probability density function with respect to Lebesgue measure on the Euclidean space R^K–1 given by

$f(x_1,\dots, x_{K-1}; \alpha_1,\dots, \alpha_K) = \frac{1}{\mathrm{B}(\alpha)} \prod_{i=1}^K x_i^{\alpha_i - 1}$

for all x₁, ..., x_K–1 > 0 satisfying x₁ + ... + x_K–1 < 1, where x_K is an abbreviation for 1 – x₁ – ... – x_K–1. The density is zero outside this open (K − 1)-dimensional simplex.

The normalizing constant is the multinomial beta function, which can be expressed in terms of the gamma function:

$\mathrm{B}(\alpha) = \frac{\prod_{i=1}^K \Gamma(\alpha_i)}{\Gamma\bigl(\sum_{i=1}^K \alpha_i\bigr)},\qquad\alpha=(\alpha_1,\dots,\alpha_K).$

Properties

Let $X = (X_1, \ldots, X_K)\sim\operatorname{Dir}(\alpha)$ , meaning that the first K – 1 components have the above density and

$X_K=1-X_1-\cdots-X_{K-1}.$

Define $\textstyle\alpha_0 = \sum_{i=1}^K\alpha_i$ . Then

$\mathrm{E}[X_i] = \frac{\alpha_i}{\alpha_0},$

$\mathrm{Var}[X_i] = \frac{\alpha_i (\alpha_0-\alpha_i)}{\alpha_0^2 (\alpha_0+1)},$

in fact, the marginals are Beta distributions:

$X_i \sim \operatorname{Beta}(\alpha_i, \alpha_0 - \alpha_i).$

Furthermore,

$\mathrm{Cov}[X_iX_j] = \frac{- \alpha_i \alpha_j}{\alpha_0^2 (\alpha_0+1)}.$

The mode of the distribution is the vector (x₁, ..., x_K) with

$x_i = \frac{\alpha_i - 1}{\alpha_0 - K}, \quad \alpha_i > 1.$

The Dirichlet distribution is conjugate to the multinomial distribution in the following sense: if

$\beta \mid X=(\beta_1, \ldots, \beta_{K}) \mid X \sim \operatorname{Mult}(X),$

where β_i is the number of occurrences of i in a sample of n points from the discrete distribution on {1, ..., K} defined by X, then

$X \mid \beta \sim \operatorname{Dir}(\alpha + \beta).$

This relationship is used in Bayesian statistics to estimate the hidden parameters, X, of a discrete probability distribution given a collection of n samples. Intuitively, if the prior is represented as Dir(α), then Dir(α + β) is the posterior following a sequence of observations with histogram β.

Entropy

If X is a Dir(α) random variable, then we can use the exponential family differential identities to get an analytic expression for the expectation of logX_i:

E[logX_i] = ψ(α_i) − ψ(α₀)

where ψ is the digamma function. This yields the following formula for the information entropy of X:

$H(X) = \log B(\alpha) + (\alpha_0-K)\psi(\alpha_0) - \sum_{j=1}^K (\alpha_j-1)\psi(\alpha_j)$

Aggregation

If $X = (X_1, \ldots, X_K)\sim\operatorname{Dir}(\alpha_1,\ldots,\alpha_K)$ , then $X' = (X_1, \ldots, X_i + X_j, \ldots, X_K)\sim\operatorname{Dir}(\alpha_1,\ldots,\alpha_i+\alpha_j,\ldots,\alpha_K)$ . This aggregation property may be used to derive the marginal distribution of X_i mentioned above.

Neutrality

(main article: neutral vector).

If $X = (X_1, \ldots, X_K)\sim\operatorname{Dir}(\alpha)$ , then the vector~X is said to be neutral^[1] in the sense that X₁ is independent of $X_2/(1-X_1),X_3/(1-X_1),\ldots,X_K/(1-X_1)$ and similarly for $X_2,\ldots,X_{K-1}$ .

Observe that any permutation of X is also neutral (a property not possessed by samples drawn from a generalized Dirichlet distribution).

Related distributions

If, for $\scriptstyle i\in\{1,2,\ldots,K\},$

$Y_i\sim\operatorname{Gamma}(\textrm{shape}=\alpha_i,\textrm{scale}=1)\text{ independently},$

then

$V=\sum_{i=1}^K Y_i\sim\operatorname{Gamma}(\textrm{shape}=\sum_{i=1}^K\alpha_i,\textrm{scale}=1),$

and

$(X_1,\ldots,X_K) = (Y_1/V,\ldots,Y_K/V)\sim \operatorname{Dir}(\alpha_1,\ldots,\alpha_K).$

Though the X_is are not independent from one another, they can be seen to be generated from a set of K independent gamma random variables. Unfortunately, since the sum V is lost in the process of forming X = (X₁, ..., X_K), it is not possible to recover the original gamma random variables from these values alone. Nevertheless, because independent random variables are simpler to work with, this reparametrization can still be useful for proofs about properties of the Dirichlet distribution.

Multinomial opinions in subjective logic are equivalent to Dirichlet distributions.

Random number generation

A method to sample a random vector $x=(x_1, \ldots, x_K)$ from the K-dimensional Dirichlet distribution with parameters $(\alpha_1, \ldots, \alpha_K)$ follows immediately from this connection. First, draw K independent random samples $y_1, \ldots, y_K$ from gamma distributions each with density

$\textrm{Gamma}(\alpha_i, 1) = \frac{y_i^{\alpha_i-1} \; e^{-y_i}}{\Gamma (\alpha_i)}, \!$

and then set

$x_i = y_i/\sum_{j=1}^K y_j. \!$

Intuitive interpretations of the parameters

String cutting

One example use of the Dirichlet distribution is if one wanted to cut strings (each of initial length 1.0) into K pieces with different lengths, where each piece had, on average, a designated average length, but allowing some variation in the relative sizes of the pieces. The α/α₀ values specify the mean lengths of the cut pieces of string resulting from the distribution. The variance around this mean varies inversely with α₀.

Pólya urn

Consider an urn containing balls of K different colors. Initially, the urn contains α₁ balls of color 1, α₂ balls of color 2, and so on. Now perform N draws from the urn, where after each draw, the ball is placed back into the urn with another ball of the same color. In the limit as N approaches infinity, the proportions of different colored balls in the urn will be distributed as $\operatorname{Dir}(\alpha_1,\ldots,\alpha_K)$ .^[2]

Note that each draw from the urn modifies the probability of drawing a ball of any one color from the urn in the future. This modification diminishes with the number of draws, since the relative effect of adding a new ball to the urn diminishes as the urn accumulates increasing numbers of balls. This "diminishing returns" effect can also help explain how large α values yield Dirichlet distributions with most of the probability mass concentrated around a single point on the simplex.

References

^ R. J. Connor and J. E. Mosiman 1969. Concepts of independence for proportions with a generalization of the Dirichlet distibution. Journal of the American Statistical Association, volume 64, pp194--206
^ D. Blackwell and J. B. MacQueen 1973. Ferguson distributions via Pólya urn schemes. The Annals of Statistics, volume 1, number 2, pp353--355

External links

v • d • e

Probability distributions

Discrete univariate with finite support

Benford · Bernoulli · binomial · categorical · hypergeometric · Rademacher · discrete uniform · Zipf · Zipf-Mandelbrot

Discrete univariate with infinite support

Boltzmann · Conway-Maxwell-Poisson · compound Poisson · discrete phase-type · extended negative binomial · Gauss-Kuzmin · geometric · logarithmic · negative binomial · parabolic fractal · Poisson · Skellam · Yule-Simon · zeta

Continuous univariate supported on a bounded interval, e.g. [0,1]

Beta · Irwin-Hall · Kumaraswamy · raised cosine · triangular · U-quadratic · uniform · Wigner semicircle

Continuous univariate supported on a semi-infinite interval, usually [0,∞)

Beta prime · Burr · chi-square · Coxian · Erlang · exponential · F · Fermi-Dirac · folded normal · Fréchet · Gamma · generalized extreme value · generalized inverse Gaussian · half-logistic · half-normal · Hotelling's T-square · hyper-exponential · hypoexponential · inverse chi-square (scale inverse chi-square) · inverse Gaussian · inverse gamma · Lévy · log-normal · log-logistic · Maxwell-Boltzmann · Maxwell speed · Nakagami · noncentral chi-square · Pareto · phase-type · Rayleigh · relativistic Breit–Wigner · Rice · Rosin–Rammler · shifted Gompertz · truncated normal · type-2 Gumbel · Weibull · Wilks' lambda

Continuous univariate supported on the whole real line (-∞,∞)

Cauchy · extreme value · exponential power · Fisher's z · Fisher-Tippett · generalized hyperbolic · hyperbolic secant · Landau · Laplace · Lévy skew alpha-stable · logistic · normal (Gaussian) · normal inverse Gaussian · skew normal · Student's t · type-1 Gumbel · Variance-Gamma · Voigt

Multivariate (joint)

Discrete: Ewens · Beta-binomial · multinomial · multivariate Polya Continuous: Dirichlet · Generalized Dirichlet · multivariate normal · multivariate Student · normal-scaled inverse gamma · normal-gamma Matrix-valued: inverse-Wishart · matrix normal · Wishart

Directional, degenerate, and singular

Directional: Kent · von Mises · von Mises–Fisher Degenerate: discrete degenerate · Dirac delta function Singular: Cantor

Families

exponential · natural exponential · location-scale · maximum entropy · Pearson · Tweedie

This entry is from Wikipedia, the leading user-contributed encyclopedia. It may not have been reviewed by professional editors (see full disclaimer)