Kumaraswamy distribution

{{Short description|Family of continuous probability distributions}}

{{Other uses|Kumaraswamy (disambiguation)}}

{{Probability distribution |

name =Kumaraswamy|

type =density|

pdf_image =Image:KumaraswamyT pdf.svg|

cdf_image =Image:Kumaraswamy cdf.svg|

parameters =a>0\, (real)
b>0\, (real)|

support =x \in (0,1)\,|

pdf =abx^{a-1}(1-x^a)^{b-1}\,|

cdf =1-(1-x^a)^b|

quantile =(1-(1-u)^\frac{1}{b})^\frac{1}{a}|

mean =\frac{b\Gamma(1+\tfrac{1}{a})\Gamma(b)}{\Gamma(1+\tfrac{1}{a}+b)}\,|

median =\left(1-2^{-1/b}\right)^{1/a}|

mode =\left(\frac{a-1}{ab-1}\right)^{1/a} for a\geq 1, b\geq 1, (a,b)\neq (1,1)|

variance =(complicated-see text)|

skewness =(complicated-see text)|

kurtosis =(complicated-see text)|

entropy =\left(1\!-\!\tfrac{1}{b}\right)+\left(1\!-\!\tfrac{1}{a}\right)H_b-\ln(ab)|

mgf =|

char =

}}

In probability and statistics, the Kumaraswamy's double bounded distribution is a family of continuous probability distributions defined on the interval (0,1). It is similar to the beta distribution, but much simpler to use especially in simulation studies since its probability density function, cumulative distribution function and quantile functions can be expressed in closed form. This distribution was originally proposed by Poondi Kumaraswamy{{Cite journal|last=Kumaraswamy|first=P.|date=1980|title=A generalized probability density function for double-bounded random processes|journal=Journal of Hydrology|volume=46|issue=1–2|pages=79–88|doi=10.1016/0022-1694(80)90036-0|issn=0022-1694|bibcode=1980JHyd...46...79K}} for variables that are lower and upper bounded with a zero-inflation. In this first article of the distribution, the natural lower bound of zero for rainfall was modelled using a discrete probability, as rainfall in many places, especially in tropics, has significant nonzero probability. This discrete probability is now called zero-inflation. This was extended to inflations at both extremes [0,1] in the work of Fletcher and Ponnambalam.{{Cite journal|last1=Fletcher|first1=S.G.|last2=Ponnambalam|first2=K.|date=1996|title=Estimation of reservoir yield and storage distribution using moments analysis|journal=Journal of Hydrology|volume=182|issue=1–4|pages=259–275|doi=10.1016/0022-1694(95)02946-x|issn=0022-1694|bibcode=1996JHyd..182..259F}} A good example for inflations at extremes are the probabilities of full and empty reservoirs and are important for reservoir design.

Characterization

=Probability density function=

The probability density function of the Kumaraswamy distribution without considering any inflation is

: f(x; a,b) = a b x^{a-1}{ (1-x^a)}^{b-1}, \ \ \mbox{where} \ \ x \in (0,1),

and where a and b are non-negative shape parameters.

=Cumulative distribution function=

The cumulative distribution function is

:F(x; a,b)=\int_{0}^{x}f(\xi; a,b) d\xi = 1-(1-x^a)^b.\

=Quantile function=

The inverse cumulative distribution function (quantile function) is

:F^{-1}(y; a,b)=(1-(1-y)^\frac{1}{b})^\frac{1}{a}. \

=Generalizing to arbitrary interval support=

In its simplest form, the distribution has a support of (0,1). In a more general form, the normalized variable x is replaced with the unshifted and unscaled variable z where:

: x = \frac{z-z_{\text{min}}}{z_{\text{max}}-z_{\text{min}}} , \qquad z_{\text{min}} \le z \le z_{\text{max}}. \,\!

Properties

The raw moments of the Kumaraswamy distribution are given by:{{Cite journal|last=Lemonte|first=Artur J.|date=2011|title=Improved point estimation for the Kumaraswamy distribution|journal=Journal of Statistical Computation and Simulation|volume=81|issue=12|pages=1971–1982|doi=10.1080/00949655.2010.511621|issn=0094-9655}}{{Cite journal|last1=CRIBARI-NETO|first1=FRANCISCO|last2=SANTOS|first2=JÉSSICA|date=2019|title=Inflated Kumaraswamy distributions|journal=Anais da Academia Brasileira de Ciências|volume=91|issue=2|pages=e20180955|doi=10.1590/0001-3765201920180955|pmid=31141016|issn=1678-2690|doi-access=free|s2cid=169034252 |url=http://pdfs.semanticscholar.org/74cf/6193ee129558fb0f7d03835852e1cdd72452.pdf}}

:m_n = \frac{b\Gamma(1+n/a)\Gamma(b)}{\Gamma(1+b+n/a)} = bB(1+n/a,b)\,

where B is the Beta function and Γ(.) denotes the Gamma function. The variance, skewness, and excess kurtosis can be calculated from these raw moments. For example, the variance is:

:\sigma^2=m_2-m_1^2.

The Shannon entropy (in nats) of the distribution is:{{cite book |last1=Michalowicz |first1=Joseph Victor |last2=Nichols |first2=Jonathan M. |last3=Bucholtz |first3=Frank |title=Handbook of Differential Entropy |date=2013 |publisher=Chapman and Hall/CRC |isbn=9781466583177 |page=100}}

:H=\left(1\!-\!\tfrac{1}{a}\right)+\left(1\!-\!\tfrac{1}{b}\right)H_b-\ln(ab)

where H_i is the harmonic number function.

Relation to the Beta distribution

The Kumaraswamy distribution is closely related to Beta distribution.{{Cite journal|last=Jones|first=M.C.|date=2009|title=Kumaraswamy's distribution: A beta-type distribution with some tractability advantages|journal=Statistical Methodology|volume=6|issue=1|pages=70–81|doi=10.1016/j.stamet.2008.04.001|issn=1572-3127}}

Assume that Xa,b is a Kumaraswamy distributed random variable with parameters a and b.

Then Xa,b is the a-th root of a suitably defined Beta distributed random variable.

More formally, Let Y1,b denote a Beta distributed random variable with parameters \alpha=1 and \beta=b.

One has the following relation between Xa,b and Y1,b.

:X_{a,b}=Y^{1/a}_{1,b},

with equality in distribution.

:\operatorname{P}\{X_{a,b}\le x\}=\int_0^x ab t^{a-1}(1-t^a)^{b-1}dt=

\int_0^{x^a} b(1-t)^{b-1}dt=\operatorname{P}\{Y_{1,b}\le x^a\}

=\operatorname{P}\{Y^{1/a}_{1,b}\le x\}

.

One may introduce generalised Kumaraswamy distributions by considering random variables of the form

Y^{1/\gamma}_{\alpha,\beta}, with \gamma>0 and where Y_{\alpha,\beta}

denotes a Beta distributed random variable with parameters \alpha and \beta.

The raw moments of this generalized Kumaraswamy distribution are given by:

:m_n = \frac{\Gamma(\alpha+\beta)\Gamma(\alpha+n/\gamma)}{\Gamma(\alpha)\Gamma(\alpha+\beta+n/\gamma)}.

Note that we can re-obtain the original moments setting \alpha=1, \beta=b and \gamma=a.

However, in general, the cumulative distribution function does not have a closed form solution.

Related distributions

  • If X \sim \textrm{Kumaraswamy}(1,1)\, then X \sim U(0,1)\, (Uniform distribution)
  • If X \sim U(0,1) \, then {\left( 1-X^{\tfrac{1}{b}} \right)}^{ \tfrac{1}{a}} \sim \textrm{Kumaraswamy}(a,b) \,
  • If X \sim \textrm{Beta}(1,b) \, (Beta distribution) then X \sim \textrm{Kumaraswamy}(1,b)\,
  • If X \sim \textrm{Beta}(a,1) \, (Beta distribution) then X \sim \textrm{Kumaraswamy}(a,1)\,
  • If X \sim \textrm{Kumaraswamy}(a,1)\, then (1-X) \sim \textrm{Kumaraswamy}(1, a)\,
  • If X \sim \textrm{Kumaraswamy}(1,a)\, then (1-X) \sim \textrm{Kumaraswamy}(a, 1)\,
  • If X \sim \textrm{Kumaraswamy}(a,1)\, then -\log(X) \sim \textrm{Exponential}(a)\,
  • If X \sim \textrm{Kumaraswamy}(1,b)\, then -\log(1-X) \sim \textrm{Exponential}(b)\,
  • If X \sim \textrm{Kumaraswamy}(a,b)\, then X \sim \textrm{GB1}(a, 1, 1, b)\,, the generalized beta distribution of the first kind
  • If X \sim \textrm{Kumaraswamy}(a,b)\, then \left\{\frac{1}{1-X} \right\} \sim \textrm{MK}(a, b)\,, the Modified Kumaraswamy distribution.

Example

An example of the use of the Kumaraswamy distribution is the storage volume of a reservoir of capacity z whose upper bound is zmax and lower bound is 0, which is also a natural example for having two inflations as many reservoirs have nonzero probabilities for both empty and full reservoir states.

References

{{Reflist}}
{{ProbDistributions|continuous-bounded}}

Category:Continuous distributions