This preview shows page 1. Sign up to view the full content.
Unformatted text preview: d, then F (b, d) − F (b, c) − F (a, d) + F (a, c) ≥ 0
• limu→−∞ F (u, v ) = 0 for each v , and limv→−∞ F (u, v ) = 0 for each u
• limu→∞ limv→∞ F (u, v ) = 1
It can be shown that any function F satisfying the above properties is the joint CDF for some pair
of random variables (X, Y ). 4.2 Joint probability mass functions If X and Y are each discretetype random variables on the same probability space, they have a
joint probability mass function (joint pmf), denoted by pX,Y . It is deﬁned by pX,Y (u, v ) = P {X =
u, Y = v }. If the numbers of possible values are small, the joint pmf can be easily described in a
table. The joint pmf determines the probabilities of any events that can be expressed as conditions
on X and Y. In particular, the pmfs of X and Y can be recovered from the joint pmf, using the
law of total probability:
pX (u) = pX,Y (u, v )
v pY (v ) = pX,Y (u, v ).
u In this case, pX and pY are called the marginal pmfs of the joint pmf, pX,Y . The conditional pmfs
are also determined by the joint pmf. For example, the conditi...
View
Full
Document
This note was uploaded on 02/09/2014 for the course ISYE 2027 taught by Professor Zahrn during the Spring '08 term at Georgia Institute of Technology.
 Spring '08
 Zahrn
 The Land

Click to edit the document details