exponential distribution
   HOME

TheInfoList



In
probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by expressing it through a set of ...
and
statistics Statistics is the discipline that concerns the collection, organization, analysis, interpretation, and presentation of data. In applying statistics to a scientific, industrial, or social problem, it is conventional to begin with a statistical ...

statistics
, the exponential distribution is the
probability distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by expre ...
of the time between events in a
Poisson point process In probability Probability is the branch of mathematics Mathematics (from Ancient Greek, Greek: ) includes the study of such topics as quantity (number theory), mathematical structure, structure (algebra), space (geometry), and calculus ...
, i.e., a process in which events occur continuously and independently at a constant average rate. It is a particular case of the
gamma distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by express ...

gamma distribution
. It is the continuous analogue of the
geometric distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by expres ...
, and it has the key property of being
memoryless In probability Probability is the branch of mathematics concerning numerical descriptions of how likely an Event (probability theory), event is to occur, or how likely it is that a proposition is true. The probability of an event is a num ...
. In addition to being used for the analysis of Poisson point processes it is found in various other contexts. The exponential distribution is not the same as the class of
exponential families In theory of probability, probability and statistics, an exponential family is a parametric model, parametric set of probability distributions of a certain form, specified below. This special form is chosen for mathematical convenience, based on ...
of distributions, which is a large class of probability distributions that includes the exponential distribution as one of its members, but also includes the
normal distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by ex ...

normal distribution
,
binomial distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by ...

binomial distribution
,
gamma distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by express ...

gamma distribution
, Poisson, and many others.


Definitions


Probability density function

The
probability density function and probability density function of a normal distribution . Image:visualisation_mode_median_mean.svg, 150px, Geometric visualisation of the mode (statistics), mode, median (statistics), median and mean (statistics), mean of an arbitrary probabilit ...
(pdf) of an exponential distribution is : f(x;\lambda) = \begin \lambda e^ & x \ge 0, \\ 0 & x < 0. \end Here ''λ'' > 0 is the parameter of the distribution, often called the ''rate parameter''. The distribution is supported on the interval . If a
random variable A random variable is a variable whose values depend on outcomes of a random In common parlance, randomness is the apparent or actual lack of pattern or predictability in events. A random sequence of events, symbols or steps often has no ...
''X'' has this distribution, we write . The exponential distribution exhibits
infinite divisibility Infinite divisibility arises in different ways in philosophy Philosophy (from , ) is the study of general and fundamental questions, such as those about reason, Metaphysics, existence, Epistemology, knowledge, Ethics, values, Philosophy o ...
.


Cumulative distribution function

The
cumulative distribution function In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by exp ...
is given by :F(x;\lambda) = \begin 1-e^ & x \ge 0, \\ 0 & x < 0. \end


Alternative parametrization

The exponential distribution is sometimes parametrized in terms of the
scale parameter In probability theory and statistics, a scale parameter is a special kind of numerical parameter of a parametric family of probability distributions. The larger the scale parameter, the more spread out the distribution. Definition If a family of p ...
, which is also the mean: f(x;\beta) = \begin \frac e^ & x \ge 0, \\ 0 & x < 0. \end \qquad\qquad F(x;\beta) = \begin 1- e^ & x \ge 0, \\ 0 & x < 0. \end


Properties


Mean, variance, moments, and median

The mean or
expected value In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by exp ...
of an exponentially distributed random variable ''X'' with rate parameter ''λ'' is given by \operatorname = \frac. In light of the examples given
below Below may refer to: *Earth *Ground (disambiguation) *Soil *Floor *Bottom (disambiguation) *Less than *Temperatures below freezing *Hell or underworld People with the surname *Fred Below (1926–1988), American blues drummer *Fritz von Below (1853 ...
, this makes sense: if you receive phone calls at an average rate of 2 per hour, then you can expect to wait half an hour for every call. The variance of ''X'' is given by \operatorname = \frac, so the standard deviation is equal to the mean. The Moment (mathematics), moments of ''X'', for n\in\N are given by \operatorname\left[X^n\right] = \frac. The central moments of ''X'', for n\in\N are given by \mu_n = \frac = \frac\sum^n_\frac. where !''n'' is the Derangement, subfactorial of ''n'' The median of ''X'' is given by \operatorname = \frac < \operatorname[X], where refers to the natural logarithm. Thus the absolute difference between the mean and median is \left, \operatorname\left[X\right] - \operatorname\left[X\right]\ = \frac < \frac = \operatorname[X], in accordance with the Chebyshev's inequality#An application: distance between the mean and the median, median-mean inequality.


Memorylessness

An exponentially distributed random variable ''T'' obeys the relation \Pr \left (T > s + t \mid T > s \right ) = \Pr(T > t), \qquad \forall s, t \ge 0. This can be seen by considering the Cumulative distribution function#Complementary cumulative distribution function (tail distribution), complementary cumulative distribution function: \begin \Pr\left(T > s + t \mid T > s\right) &= \frac \\[4pt] &= \frac \\[4pt] &= \frac \\[4pt] &= e^ \\[4pt] &= \Pr(T > t). \end When ''T'' is interpreted as the waiting time for an event to occur relative to some initial time, this relation implies that, if ''T'' is conditioned on a failure to observe the event over some initial period of time ''s'', the distribution of the remaining waiting time is the same as the original unconditional distribution. For example, if an event has not occurred after 30 seconds, the conditional probability that occurrence will take at least 10 more seconds is equal to the unconditional probability of observing the event more than 10 seconds after the initial time. The exponential distribution and the
geometric distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by expres ...
are memorylessness, the only memoryless probability distributions. The exponential distribution is consequently also necessarily the only continuous probability distribution that has a constant failure rate.


Quantiles

The quantile function (inverse cumulative distribution function) for Exp(''λ'') is F^(p;\lambda) = \frac,\qquad 0 \le p < 1 The quartiles are therefore: *first quartile: ln(4/3)/''λ'' *median: ln(2)/''λ'' *third quartile: ln(4)/''λ'' And as a consequence the interquartile range is ln(3)/''λ''.


Kullback–Leibler divergence

The directed Kullback–Leibler divergence in nat (unit), nats of e^\lambda ("approximating" distribution) from e^ ('true' distribution) is given by \begin \Delta(\lambda_0 \parallel \lambda) &= \mathbb_\left( \log \frac\right)\\ &= \mathbb_\left( \log \frac\right)\\ &= \log(\lambda_0) - \log(\lambda) - (\lambda_0 - \lambda)E_(x)\\ &= \log(\lambda_0) - \log(\lambda) + \frac - 1. \end


Maximum entropy distribution

Among all continuous probability distributions with Support (mathematics)#In probability and measure theory, support and mean ''μ'', the exponential distribution with ''λ'' = 1/''μ'' has the largest differential entropy. In other words, it is the maximum entropy probability distribution for a random variate ''X'' which is greater than or equal to zero and for which E[''X''] is fixed.


Distribution of the minimum of exponential random variables

Let ''X''1, …, ''X''''n'' be Independent random variables, independent exponentially distributed random variables with rate parameters ''λ''1, …, ''λn''. Then \min\left\ is also exponentially distributed, with parameter \lambda = \lambda_1 + \dotsb + \lambda_n. This can be seen by considering the Cumulative distribution function#Complementary cumulative distribution function (tail distribution), complementary cumulative distribution function: \begin &\Pr\left(\min\ > x\right) \\ = &\Pr\left(X_1 > x, \dotsc, X_n > x\right) \\ = &\prod_^n \Pr\left(X_i > x\right) \\ = &\prod_^n \exp\left(-x\lambda_i\right) = \exp\left(-x\sum_^n \lambda_i\right). \end The index of the variable which achieves the minimum is distributed according to the categorical distribution \Pr\left(k \mid X_k = \min\\right) = \frac. A proof can be seen by letting I = \operatorname_\. Then, \begin \Pr (I = k) &= \int_^ \Pr(X_k = x) \Pr(X_ > x ) \,dx \\ &= \int_^ \lambda_k e^ \left(\prod_^ e^\right) dx \\ &= \lambda_k \int_^ e^ dx \\ &= \frac. \end Note that \max\ is not exponentially distributed, if ''X''1, …, ''X''''n'' do not all have parameter 0.


Joint moments of i.i.d. exponential order statistics

Let X_1, \dotsc, X_n be n Independent and identically distributed random variables, independent and identically distributed exponential random variables with rate parameter ''λ''. Let X_, \dotsc, X_ denote the corresponding order statistics. For i < j , the joint moment \operatorname E\left[X_ X_\right] of the order statistics X_ and X_ is given by \begin \operatorname E\left[X_ X_\right] &= \sum_^\frac \operatorname E\left[X_\right] + \operatorname E\left[X_^2\right] \\ &= \sum_^\frac\sum_^\frac + \sum_^\frac + \left(\sum_^\frac\right)^2. \end This can be seen by invoking the law of total expectation and the memoryless property: \begin \operatorname E\left[X_ X_\right] &= \int_0^\infty \operatorname E\left[X_ X_ \mid X_=x\right] f_(x) \, dx \\ &= \int_^\infty x \operatorname E\left[X_ \mid X_ \geq x\right] f_(x) \, dx &&\left(\textrm~X_ = x \implies X_ \geq x\right) \\ &= \int_^\infty x \left[ \operatorname E\left[X_\right] + x \right] f_(x) \, dx &&\left(\text\right) \\ &= \sum_^\frac \operatorname E\left[X_\right] + \operatorname E\left[X_^2\right]. \end The first equation follows from the law of total expectation. The second equation exploits the fact that once we condition on X_ = x , it must follow that X_ \geq x . The third equation relies on the memoryless property to replace \operatorname E\left[ X_ \mid X_ \geq x\right] with \operatorname E\left[X_\right] + x.


Sum of two independent exponential random variables

The probability distribution function (PDF) of a sum of two independent random variables is the convolution of probability distributions, convolution of their individual PDFs. If X_1 and X_2 are independent exponential random variables with respective rate parameters \lambda_1 and \lambda_2, then the probability density of Z=X_1+X_2 is given by \begin f_Z(z) &= \int_^\infty f_(x_1) f_(z - x_1)\,dx_1\\ &= \int_0^z \lambda_1 e^ \lambda_2 e^ \, dx_1 \\ &= \lambda_1 \lambda_2 e^ \int_0^z e^\,dx_1 \\ &= \begin \dfrac \left(e^ - e^\right) & \text \lambda_1 \neq \lambda_2 \\[4 pt] \lambda^2 z e^ & \text \lambda_1 = \lambda_2 = \lambda. \end \end The entropy of this distribution is available in closed form: assuming \lambda_1 > \lambda_2 (without loss of generality), then \begin H(Z) &= 1 + \gamma + \ln \left( \frac \right) + \psi \left( \frac \right) , \end where \gamma is the Euler-Mascheroni constant, and \psi(\cdot) is the digamma function. In the case of equal rate parameters, the result is an Erlang distribution with shape 2 and parameter \lambda, which in turn is a special case of
gamma distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by express ...

gamma distribution
.


Related distributions

* If X \sim \operatorname\left(\mu, \beta^\right) then , ''X'' − μ, ~ Exp(β). * If ''X'' ~ Pareto(1, λ) then log(''X'') ~ Exp(λ). * If ''X'' ~ skew-logistic distribution, SkewLogistic(θ), then \log\left(1 + e^\right) \sim \operatorname(\theta). * If ''Xi'' ~ Uniform distribution (continuous), ''U''(0, 1) then \lim_n \min \left(X_1, \ldots, X_n\right) \sim \operatorname(1) * The exponential distribution is a limit of a scaled beta distribution: \lim_ n \operatorname(1, n) = \operatorname(1). * Exponential distribution is a special case of type 3 Pearson distribution. * If ''X'' ~ Exp(λ) and ''X'' ~ Exp(λ) then: ** kX \sim \operatorname\left(\frac\right), closure under scaling by a positive factor. ** 1 + ''X'' ~ Benktander Weibull distribution, BenktanderWeibull(λ, 1), which reduces to a truncated exponential distribution. ** ''keX'' ~ Pareto distribution, Pareto(''k'', λ). ** ''e−X'' ~ Beta distribution, Beta(λ, 1). ** ''e'' ~ power law, PowerLaw(''k'', λ) ** \sqrt \sim \operatorname \left(\frac\right), the Rayleigh distribution ** X \sim \operatorname\left(\frac, 1\right), the Weibull distribution ** X^2 \sim \operatorname\left(\frac, \frac\right) ** . ** \lfloor X\rfloor \sim \operatorname\left(1-e^\right), a
geometric distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by expres ...
on 0,1,2,3,... ** \lceil X\rceil \sim \operatorname\left(1-e^\right), a
geometric distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by expres ...
on 1,2,3,4,... ** If also ''Y'' ~ Erlang(''n'', λ) orY \sim \Gamma\left(n, \frac\right) then \frac + 1 \sim \operatorname(1, n) ** If also λ ~ gamma distribution, Gamma(''k'', θ) (shape, scale parametrisation) then the marginal distribution of ''X'' is Lomax distribution, Lomax(''k'', 1/θ), the gamma compound distribution, mixture ** λ''X'' − λ''Y'' ~ Laplace distribution, Laplace(0, 1). ** min ~ Exp(λ1 + ... + λ''n''). ** If also λ = λ then: *** X_1 + \cdots + X_k = \sum_i X_i \sim Erlang distribution, Erlang(''k'', λ) = gamma distribution, Gamma(''k'', λ−1) = Gamma(''k'', λ) (in (''k'', θ) and (α, β) parametrization, respectively) with an integer shape parameter k. *** ''X'' − ''X'' ~ Laplace(0, λ−1). ** If also ''X'' are independent, then: *** \frac ~ uniform distribution (continuous), U(0, 1) *** Z = \frac has probability density function f_Z(z) = \frac. This can be used to obtain a confidence interval for \frac. ** If also λ = 1: *** \mu - \beta\log\left(\frac\right) \sim \operatorname(\mu, \beta), the logistic distribution *** \mu - \beta\log\left(\frac\right) \sim \operatorname(\mu, \beta) *** ''μ'' − σ log(''X'') ~ generalized extreme value distribution, GEV(μ, σ, 0). *** Further if Y \sim \Gamma\left(\alpha, \frac\right) then \sqrt \sim \operatorname(\alpha, \beta) (K-distribution) ** If also λ = 1/2 then ; i.e., ''X'' has a chi-squared distribution with 2 degrees of freedom (statistics), degrees of freedom. Hence: \operatorname(\lambda) = \frac \operatorname\left(\frac \right) \sim \frac \chi_2^2\Rightarrow \sum_^n \operatorname(\lambda) \sim \frac\chi_^2 * If X \sim \operatorname\left(\frac\right) and Y \mid X ~ Poisson distribution, Poisson(''X'') then Y \sim \operatorname\left(\frac\right) (
geometric distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by expres ...
) * The Hoyt distribution can be obtained from exponential distribution and arcsine distribution Other related distributions: *Hyper-exponential distribution – the distribution whose density is a weighted sum of exponential densities. *Hypoexponential distribution – the distribution of a general sum of exponential random variables. *exGaussian distribution – the sum of an exponential distribution and a
normal distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by ex ...

normal distribution
.


Statistical inference

Below, suppose random variable ''X'' is exponentially distributed with rate parameter λ, and x_1, \dotsc, x_n are ''n'' independent samples from ''X'', with sample mean \bar.


Parameter estimation

The maximum likelihood estimator for λ is constructed as follows: The likelihood function for λ, given an independent identically-distributed random variables, independent and identically distributed sample ''x'' = (''x''1, …, ''x''''n'') drawn from the variable, is: L(\lambda) = \prod_^n\lambda\exp(-\lambda x_i) = \lambda^n\exp\left(-\lambda \sum_^n x_i\right) = \lambda^n\exp\left(-\lambda n\overline\right), where: \overline = \frac\sum_^n x_i is the sample mean. The derivative of the likelihood function's logarithm is: \frac \ln L(\lambda) = \frac \left( n \ln\lambda - \lambda n\overline \right) = \frac - n\overline\ \begin > 0, & 0 < \lambda < \frac, \\[8pt] = 0, & \lambda = \frac, \\[8pt] < 0, & \lambda > \frac. \end Consequently, the maximum likelihood estimate for the rate parameter is: \widehat_\text = \frac = \frac This is an unbiased estimator of \lambda, although \overline an unbiased MLE estimator of 1/\lambda and the distribution mean. The bias of \widehat_\text is equal to b \equiv \operatorname\left[\left(\widehat_\text - \lambda\right)\right] = \frac which yields the Maximum likelihood estimation#Second-order efficiency after correction for bias, bias-corrected maximum likelihood estimator \widehat^*_\text = \widehat_\text - \widehat .


Approximate minimizer of expected squared error

Assume you have at least three samples. If we seek a minimizer of expected mean squared error (see also: Bias–variance tradeoff) that is similar to the maximum likelihood estimate (i.e. a multiplicative correction to the likelihood estimate) we have: \widehat = \left(\frac\right) \left(\frac\right) = \frac This is derived from the mean and variance of the inverse-gamma distribution: \mbox(n, \lambda).


Fisher information

The Fisher information, denoted \mathcal(\lambda), for an estimator of the rate parameter \lambda is given as: \mathcal(\lambda) = \operatorname \left[\left. \left(\frac \log f(x;\lambda)\right)^2\\lambda\right] = \int \left(\frac \log f(x;\lambda)\right)^2 f(x; \lambda)\,dx Plugging in the distribution and solving gives: \mathcal(\lambda) = \int_^ \left(\frac \log \lambda e^\right)^2 \lambda e^\,dx = \int_^ \left(\frac - x\right)^2 \lambda e^\,dx = \lambda^. This determines the amount of information each independent sample of an exponential distribution carries about the unknown rate parameter \lambda.


Confidence intervals

The 100(1 − α)% confidence interval for the rate parameter of an exponential distribution is given by: \frac< \frac < \frac which is also equal to: \frac < \frac < \frac where is the percentile of the chi squared distribution with ''v'' degrees of freedom (statistics), degrees of freedom, n is the number of observations of inter-arrival times in the sample, and x-bar is the sample average. A simple approximation to the exact interval endpoints can be derived using a normal approximation to the distribution. This approximation gives the following values for a 95% confidence interval: \begin \lambda_\text &= \widehat\left(1 - \frac\right) \\ \lambda_\text &= \widehat\left(1 + \frac\right) \end This approximation may be acceptable for samples containing at least 15 to 20 elements.


Bayesian inference

The conjugate prior for the exponential distribution is the
gamma distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by express ...

gamma distribution
(of which the exponential distribution is a special case). The following parameterization of the gamma probability density function is useful: \operatorname(\lambda; \alpha, \beta) = \frac \lambda^ \exp(-\lambda\beta). The posterior distribution ''p'' can then be expressed in terms of the likelihood function defined above and a gamma prior: \begin p(\lambda) &\propto L(\lambda) \Gamma(\lambda; \alpha, \beta) \\ &= \lambda^n \exp\left(-\lambda n\overline\right) \frac \lambda^ \exp(-\lambda \beta) \\ &\propto \lambda^ \exp(-\lambda \left(\beta + n\overline\right)). \end Now the posterior density ''p'' has been specified up to a missing normalizing constant. Since it has the form of a gamma pdf, this can easily be filled in, and one obtains: p(\lambda) = \Gamma(\lambda; \alpha + n, \beta + n\overline). Here the hyperparameter ''α'' can be interpreted as the number of prior observations, and ''β'' as the sum of the prior observations. The posterior mean here is: \frac.


Occurrence and applications


Occurrence of events

The exponential distribution occurs naturally when describing the lengths of the inter-arrival times in a homogeneous Poisson process. The exponential distribution may be viewed as a continuous counterpart of the
geometric distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by expres ...
, which describes the number of Bernoulli trials necessary for a ''discrete'' process to change state. In contrast, the exponential distribution describes the time for a continuous process to change state. In real-world scenarios, the assumption of a constant rate (or probability per unit time) is rarely satisfied. For example, the rate of incoming phone calls differs according to the time of day. But if we focus on a time interval during which the rate is roughly constant, such as from 2 to 4 p.m. during work days, the exponential distribution can be used as a good approximate model for the time until the next phone call arrives. Similar caveats apply to the following examples which yield approximately exponentially distributed variables: * The time until a radioactive particle decays, or the time between clicks of a Geiger counter * The time it takes before your next telephone call * The time until default (on payment to company debt holders) in reduced-form credit risk modeling Exponential variables can also be used to model situations where certain events occur with a constant probability per unit length, such as the distance between mutations on a DNA strand, or between roadkills on a given road. In queuing theory, the service times of agents in a system (e.g. how long it takes for a bank teller etc. to serve a customer) are often modeled as exponentially distributed variables. (The arrival of customers for instance is also modeled by the Poisson distribution if the arrivals are independent and distributed identically.) The length of a process that can be thought of as a sequence of several independent tasks follows the Erlang distribution (which is the distribution of the sum of several independent exponentially distributed variables). Reliability theory and reliability engineering also make extensive use of the exponential distribution. Because of the ''#Memorylessness, memoryless'' property of this distribution, it is well-suited to model the constant hazard rate portion of the bathtub curve used in reliability theory. It is also very convenient because it is so easy to add failure rates in a reliability model. The exponential distribution is however not appropriate to model the overall lifetime of organisms or technical devices, because the "failure rates" here are not constant: more failures occur for very young and for very old systems. In physics, if you observe a gas at a fixed temperature and pressure in a uniform gravitational field, the heights of the various molecules also follow an approximate exponential distribution, known as the Barometric formula. This is a consequence of the entropy property mentioned below. In hydrology, the exponential distribution is used to analyze extreme values of such variables as monthly and annual maximum values of daily rainfall and river discharge volumes. :The blue picture illustrates an example of fitting the exponential distribution to ranked annually maximum one-day rainfalls showing also the 90% confidence belt based on the
binomial distribution In probability theory Probability theory is the branch of mathematics concerned with probability. Although there are several different probability interpretations, probability theory treats the concept in a rigorous mathematical manner by ...

binomial distribution
. The rainfall data are represented by plotting positions as part of the cumulative frequency analysis. In operating-rooms management, the distribution of surgery duration for a category of surgeries with Predictive methods for surgery duration, no typical work-contnet (like in an emergency room, encompassing all types of surgeries).


Prediction

Having observed a sample of ''n'' data points from an unknown exponential distribution a common task is to use these samples to make predictions about future data from the same source. A common predictive distribution over future samples is the so-called plug-in distribution, formed by plugging a suitable estimate for the rate parameter ''λ'' into the exponential density function. A common choice of estimate is the one provided by the principle of maximum likelihood, and using this yields the predictive density over a future sample ''x''''n''+1, conditioned on the observed samples ''x'' = (''x''1, ..., ''xn'') given by p_(x_ \mid x_1, \ldots, x_n) = \left( \frac1 \right) \exp \left( - \frac \right) The Bayesian approach provides a predictive distribution which takes into account the uncertainty of the estimated parameter, although this may depend crucially on the choice of prior. A predictive distribution free of the issues of choosing priors that arise under the subjective Bayesian approach is p_(x_ \mid x_1, \ldots, x_n) = \frac, which can be considered as # a frequentist confidence distribution, obtained from the distribution of the pivotal quantity /; # a profile predictive likelihood, obtained by eliminating the parameter ''λ'' from the joint likelihood of ''x''''n''+1 and ''λ'' by maximization; # an objective Bayesian predictive posterior distribution, obtained using the non-informative Jeffreys prior 1/''λ''; # the Conditional Normalized Maximum Likelihood (CNML) predictive distribution, from information theoretic considerations. The accuracy of a predictive distribution may be measured using the distance or divergence between the true exponential distribution with rate parameter, ''λ''0, and the predictive distribution based on the sample ''x''. The Kullback–Leibler divergence is a commonly used, parameterisation free measure of the difference between two distributions. Letting Δ(''λ''0, , ''p'') denote the Kullback–Leibler divergence between an exponential with rate parameter ''λ''0 and a predictive distribution ''p'' it can be shown that \begin \operatorname_ \left[ \Delta(\lambda_0\parallel p_) \right] &= \psi(n) + \frac - \log(n) \\ \operatorname_ \left[ \Delta(\lambda_0\parallel p_) \right] &= \psi(n) + \frac - \log(n) \end where the expectation is taken with respect to the exponential distribution with rate parameter , and is the digamma function. It is clear that the CNML predictive distribution is strictly superior to the maximum likelihood plug-in distribution in terms of average Kullback–Leibler divergence for all sample sizes .


Computational methods


Generating exponential variates

A conceptually very simple method for generating exponential random variate, variates is based on inverse transform sampling method, inverse transform sampling: Given a random variate ''U'' drawn from the uniform distribution (continuous), uniform distribution on the unit interval , the variate T = F^(U) has an exponential distribution, where ''F'' is the quantile function, defined by F^(p)=\frac. Moreover, if ''U'' is uniform on (0, 1), then so is 1 − ''U''. This means one can generate exponential variates as follows: T = \frac. Other methods for generating exponential variates are discussed by KnuthDonald Knuth, Donald E. Knuth (1998). ''The Art of Computer Programming'', volume 2: ''Seminumerical Algorithms'', 3rd edn. Boston: Addison–Wesley. . ''See section 3.4.1, p. 133.'' and Devroye.Luc Devroye (1986).
Non-Uniform Random Variate Generation
'. New York: Springer-Verlag. . ''Se
chapter IX
section 2, pp. 392–401.''
A fast method for generating a set of ready-ordered exponential variates without using a sorting routine is also available.


See also

* Dead time – an application of exponential distribution to particle detector analysis. * Laplace distribution, or the "double exponential distribution". * Relationships among probability distributions * Marshall–Olkin exponential distribution


References


External links

*
Online calculator of Exponential Distribution
{{DEFAULTSORT:Exponential Distribution Continuous distributions Exponentials Poisson point processes Conjugate prior distributions Exponential family distributions Infinitely divisible probability distributions Survival analysis