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 random variables, density functions, and measures September 29, 2010 1 Probability density functions and probabil ity measures We defined a probability density function (pdf for short) as a function f : R [0 , ) satisfying integraldisplay R f ( x ) dx = 1 . (Of course, this means that f must be Lebesgueintegrable, and hence mea surable.) Likewise, we defined a probability mass function, which I will also refer to as a pdf, as a function p : { a 1 , a 2 , a 3 , ... } [0 , 1] satisfying summationdisplay i p ( a i ) = 1 . So, mass functions are analogues of density functions for discrete (finite, or possibly countably infinite) sets of real numbers. What a pdf gives us is a probability measure for subsets of the real num bers. Specifically, suppose A B ( R ). Then, we can define P ( A ) := integraldisplay R 1 A ( x ) f ( x ) dx = integraldisplay A f ( x ) dx. (1) And we can do the same in the discrete case, for if A { a 1 , a 2 , ... } , then we can define P ( A ) := summationdisplay a i A p ( a i ) . 1 There is a way to unify both of these types of measures: Basically, let : B ( R ) [0 , ). Then, for A B ( R ) we use the notation integraldisplay R 1 A ( x ) d ( x ) , or , alternatively...
View Full
Document
 Spring '08
 Staff
 Statistics, Probability

Click to edit the document details