Beta distribution
The beta distribution is a continuous probability distribution over the interval.
- 2.1 extremum
- 2.2 Expectation value
- 2.3 variance
- 2.4 standard deviation
- 2.5 Coefficient of variation
- 2.6 skewness
- 2.7 symmetry
- 3.1 Relationship to F-distribution
- 3.2 Relationship to the gamma distribution
Definition
Beta distribution on [0,1]
Beta distribution is defined by the probability density
Outside of the interval it is continued by. It has the real parameter and. In order to guarantee their normalizability, is required.
The pre-factor is used to correct normalization. The term
Represents the Beta function, according to which the distribution is designated. This refers to the gamma function.
The distribution function is correspondingly
This function is also called regularized incomplete beta function.
Beta distribution on [a, b]
The general beta distribution is defined to
Wherein a and b are the lower and upper limits of the interval. Accordingly, the calculation of results to
The other models in this article applies only to the restricted to the interval Beta distribution.
Properties
Extremum
The density function takes its extremum at the position.
Expected value
The expected value is calculated as
Variance
The variance is given by
Standard deviation
For the standard deviation results
Coefficient of variation
From the expected value and variance are obtained directly the coefficient of variation
Skew
The skewness is given by
Symmetry
The beta distribution is symmetric about the skewness.
Relations with other distributions
Relationship to the F-distribution
When is F- distributed and then distributed
Relationship to the gamma distribution
If and independent gamma distributed random variables with parameters and, then the size is beta distributed with parameters and, shortly
Example
The beta distribution can be obtained from two gamma distributions: The quotient of the stochastically independent random variables, both of which are gamma distributed with parameters and respectively, is beta distributed with parameters and. and can be interpreted as a chi-square distribution with degrees of freedom respectively.
By means of linear regression, a regression line is placed by a point cloud with two pairs of values , and statistical features, and such that the sum of squares of the vertical distances of the levels of the line are minimized.
The total scattering of y ( TSS) can be decomposed into the so-called scatter explained by a straight line estimated values y * (ESS ) and the unexplained scatter of the residuals ( RSS ) with the scattering decomposition:
The coefficient of determination, the proportion of explained dispersion of the total scattering
Respectively
So is beta distributed. Since the coefficient of determination is the square of the correlation coefficient and also the square of the correlation coefficient is beta distributed.
However, the distribution of the coefficient of determination are specified in the model test of the regression through the F-distribution, which is present tabulated.