AP Statistics Curriculum 2007 Pareto

From SOCR
Revision as of 16:24, 11 July 2011 by TracyTam (talk | contribs) (Pareto Distribution)
Jump to: navigation, search

Pareto Distribution

Definition: Pareto distribution is a skewed, heavy-tailed distribution that is sometimes used to model that distribution of incomes. The basis of the distribution is that a high proportion of a population has low income while only a few people have very high incomes.


Probability density function: For \(X\sim \operatorname{Pareto}(x_m,\alpha)\!\), the Pareto probability density function is given by

\[\frac{\alpha x_m^\alpha}{x^{\alpha+1}}\]

where

  • \(x_m\) is the minimum possible value of X
  • \(\alpha\) is a positive parameter which determines the concentration of data towards the mode
  • x is a random variable (\(x>x_m\))


Cumulative density function: The Pareto cumulative distribution function is given by

\[1-(\frac{x_m}{x})^\alpha\]

where

  • \(x_m\) is the minimum possible value of X
  • \(\alpha\) is a positive parameter which determines the concentration of data towards the mode
  • x is a random variable (\(x>x_m\))


Moment generating function: The Pareto moment-generating function is

\[M(t)=\alpha(-x_m t)^\alpha\Gamma(-\alpha,-x_m t)\!\]

where

  • \(\textstyle\Gamma(-\alpha,-x_m t)=\int_{-x_m t}^\infty t^{-\alpha-1}e^{-t}dt\)


Expectation: The expected value of Pareto distributed random variable x is

\[E(X)=\frac{\alpha x_m}{\alpha-1}\mbox{ for }\alpha>1\!\]


Variance: The Pareto variance is

\[Var(X)=\frac{x_m^2 \alpha}{(\alpha-1)^2(\alpha-2)}\mbox{ for }\alpha>2\!\]

Applications

The Pareto distribution is sometimes expressed more simply as the “80-20 rule”, which describes a range of situations. In customer support, it means that 80% of problems come from 20% of customers. In economics, it means 80% of the wealth is controlled by 20% of the population. Examples of events that may be modeled by Pareto distribution include:

  • The sizes of human settlements (few cities, many villages)
  • The file size distribution of Internet traffic which uses the TCP protocol (few larger files, many smaller files)
  • Hard disk drive error rates
  • The values of oil reserves in oil fields (few large fields, many small fields)
  • The length distribution in jobs assigned supercomputers (few large ones, many small ones)
  • The standardized price returns on individual stocks
  • The sizes of sand particles
  • The sizes of meteorites
  • The number of species per genus
  • The areas burned in forest fires
  • The severity of large casualty losses for certain businesses, such as general liability, commercial auto, and workers compensation

Example

Suppose that the income of a certain population has a Pareto distribution with \(\alpha=3\) and \(x_m=1000\). Compute the proportion of the population with incomes between 2000 and 4000.

We can compute this as follows:

\[P(2000\le X\le 4000)=\sum_{x=2000}^{4000}\frac{3\times 1000^3}{x^{3+1}}=0.109375\]

The figure below shows this result using SOCR distributions

Pareto.jpg