Difference between revisions of "Gamma distribution"

From timescalewiki
Jump to: navigation, search
Line 1: Line 1:
Let $\mathbb{T}$ be a time scale. Let $\lambda \in \mathbb{R}$ with $\lambda > 0$ and define $\Lambda_0(t,t_0)=0, \Lambda_1(t,t_0)=1$. The gamma distribution is the [[probability density function]] defined recursively for $k \geq 2$ by the formula
+
Let $\mathbb{T}$ be a time scale. Let $\lambda \in \mathbb{R}$ with $\lambda > 0$ and define $\Lambda_0(t,t_0)=0, \Lambda_1(t,t_0)=1$ and
 
$$\Lambda_{k+1}(t,t_0) = -\displaystyle\int_{t_0}^t (\ominus \lambda)(\tau) \Lambda_k(\sigma(\tau),t_0) \Delta \tau.$$
 
$$\Lambda_{k+1}(t,t_0) = -\displaystyle\int_{t_0}^t (\ominus \lambda)(\tau) \Lambda_k(\sigma(\tau),t_0) \Delta \tau.$$
 +
 +
The gamma [[probability density function]] is defined to be
 +
$$f_k(t)= \left\{ \begin{array}{ll}
 +
\dfrac{\lambda}{e_{\lambda}(\sigma(t),0)} \Lambda_k(\sigma(t),0) &; t \geq 0 \\
 +
0 &; t < 0
 +
\end{array} \right.$$
 +
 +
=Properties=
 +
{{:Expected value of gamma distribution}}
 +
{{:Variance of gamma distribution}}
  
 
{{:Probability distributions footer}}
 
{{:Probability distributions footer}}

Revision as of 22:13, 14 April 2015

Let $\mathbb{T}$ be a time scale. Let $\lambda \in \mathbb{R}$ with $\lambda > 0$ and define $\Lambda_0(t,t_0)=0, \Lambda_1(t,t_0)=1$ and $$\Lambda_{k+1}(t,t_0) = -\displaystyle\int_{t_0}^t (\ominus \lambda)(\tau) \Lambda_k(\sigma(\tau),t_0) \Delta \tau.$$

The gamma probability density function is defined to be $$f_k(t)= \left\{ \begin{array}{ll} \dfrac{\lambda}{e_{\lambda}(\sigma(t),0)} \Lambda_k(\sigma(t),0) &; t \geq 0 \\ 0 &; t < 0 \end{array} \right.$$

Properties

Theorem: Let $X$ have the gamma distribution on $\mathbb{T}$. Then $$\mathrm{E}_{\mathbb{T}}(X)=\dfrac{k}{\lambda}.$$

Proof:

Theorem: Let $X$ have the gamma distribution on $\mathbb{T}$. Then $$\mathrm{Var}_{\mathbb{T}}(X)=\dfrac{k}{\lambda^2}.$$

Proof:

Probability distributions

Uniform distributionExponential distributionGamma distribution