This preview shows pages 1–3. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Notes on Bernoulli and Binomial random variables October 1, 2010 1 Expectation and Variance 1.1 Definitions I suppose it is a good time to talk about expectation and variance, since they will be needed in our discussion on Bernoulli and Binomial random variables, as well as for later disucssion (in a forthcoming lecture) of Poisson processes and Poisson random variables. Basically, given a random variable X : S → R , having a pdf f ( x ), define the expectation to be E ( X ) := integraldisplay ∞ −∞ xf ( x ) dx. In other words, it is kind of like the “average value of X weighted by f ( x )”. If X is discrete, where say X can take on any of the values a 1 , a 2 , ... , we would have E ( X ) := summationdisplay i a i p ( a i ) , where p here denotes the mass function. More generally, if ν : B ( R ) → [0 , 1] is a probability measure associated to a random variable X (continuous, discrete, or otherwise), so that for A ∈ B ( R ) we have P ( X ∈ A ) = integraldisplay R dν ( x ) = ν ( A ) , 1 we define E ( X ) := integraldisplay R xdν ( x ) . Of course, we really haven’t worked much with the sort of r.v.’s for which such a ν cannot be relized in terms of pdf’s, but I thought I would point it out anyways. We likewise define the variance operator V ( X ) to be V ( X ) := integraldisplay R ( x − μ ) 2 f ( x ) dx, where μ = E ( X ) . If X is discrete, then V ( X ) := summationdisplay i ( a i − μ ) 2 p ( a i ); and, of course, there is the even more general version V ( X ) := integraldisplay R ( x − μ ) 2 dμ ( x ) . We know that the expectation is a kind of average, and now I want to give you a feel for what the variance is: Basically, it is a measure of how “flat” the pdf is – the flatter it is, the more the values of X away from the mean μ get weighted; and therefore the larger the variance V ( X ) is. For example, consider the two random variables X and Y having pdf’s f ( x ) and g ( x ), respectively, given by f ( x ) = braceleftbigg 1 / 2 , if x ∈ [ − 1 , 1]; , if x < − 1 or x > 1 ....
View
Full
Document
This note was uploaded on 10/23/2011 for the course MATH 3225 taught by Professor Staff during the Spring '08 term at Georgia Institute of Technology.
 Spring '08
 Staff
 Statistics, Bernoulli, Binomial, Probability, Variance

Click to edit the document details