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: 1.2. CONDITIONAL EXPECTATION 13 1.2 Conditional Expectation Let Λ ∈ F with P (Λ) > 0, we define P ( E  Λ) = P (Λ ∩ E ) P (Λ) where P (Λ) > . It follow that for a random vector ( X,Y ), P ( Y ≤ y  X = x ) = P ( Y ≤ y, X = x ) P ( X = x ) , if P ( X = x ) > , , otherwise . Moreover if ( X,Y ) has joint density f ( x,y ), the conditional density of Y given X = x is f ( y  x ) = f ( x,y ) f X ( x ) , if f X ( x ) > , , otherwise . where f X ( x ) = R ∞∞ f ( x,y ) dy is the marginal density. The conditional expec tation of Y given X = x is E ( Y  X = x ) = Z ∞∞ yf ( y  x ) dy. Note that g ( x ) := E ( Y  X = x ) is a function on x , and hence g ( X ( · )) := E ( Y  X ( · )) is a r.v. on Ω . (1.2.1) In the following we have a more general consideration for the conditional expectation (and also the conditional probability): E ( Y G ) where G is a sub σfield of F . First let us look at a special case where G is generated by a measurable partition { Λ n } n of Ω (each member in G is a union of { Λ n } n ). Let Y be an 14 CHAPTER 1. BASIC PROBABILITY THEORY integrable r.v., then E ( Y  Λ n ) = Z Ω Y ( ω ) dP Λ n ( ω ) = 1 P (Λ n ) Z Λ n Y ( ω ) dP ( ω ) . (1.2.2) (Here P Λ n ( · ) = P ( · ∩ Λ n ) P (Λ n ) is a probability measure for P (Λ n ) > 0). Consider the random variable (as in (1.2.1)) Z ( · ) = E ( Y G )( · ) := X n E ( Y  Λ n ) χ Λ n ( · ) ∈ G . It is easy to see that if ω ∈ Λ n , then Z ( ω ) = E ( Y  Λ n ), and moreover Z Ω E ( Y G ) dP = X n Z Λ n E ( Y G ) dP = X n E ( Y  Λ n ) P (Λ n ) = Z Ω Y dP . We can also replace Ω by Λ ∈ G and obtain Z Λ E ( Y G ) dP = Z Λ Y dP ∀ Λ ∈ G ....
View
Full
Document
This note was uploaded on 05/23/2010 for the course MATH 987 taught by Professor Cheung,cecilia during the Spring '08 term at CUHK.
 Spring '08
 CHEUNG,CECILIA
 Calculus

Click to edit the document details