This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Expectations Enrique CamposN´ a˜nez The George Washington University ApSc 116 1 / 41 Table of Contents The Expectation of a Random Variable Expectations and Joint Distributions Properties of the Expectations Expectation of Nonnegative r.v.’s Variance Moments Covariance and Correlation Conditional Expectation Probability Theorems 2 / 41 Definition (Expectation) Let X be a discrete r.v. with p.f. f ( x ). We define the expectation of X as E ( X ) ≡ X x xf ( x ) , also called the mean of X . We say the expectation exists if X x  x  f ( x ) < ∞ . Example (A Single Die) Let X be the result of a single die toss. Then E ( X ) = X x xf ( x ) = 6 X x =1 x 1 6 = 1 6 [1 + 2 + 3 + 4 + 5 + 6] = 3 . 5 . 3 / 41 Example (Geometric Random Variable) Let 0 < p < 1, and let X be a random variable representing the number of attempts necessary to obtain a success in a Beroulli trial with success probability p . Then X is a geometric r.v., and its p.f. is f ( x ) = (1 p ) x 1  {z } x 1 failures p {z} success , x = 1 , 2 ,... We now compute its expected value or mean, i.e., E ( x ) = ∞ X x =1 x (1 p ) x 1 p = 1 p Note: this result requires the use of a moment generating function, which we will cover later in the semester. 4 / 41 Definition (Expectation of a Continuous r.v.) Let X be a continuous r.v. with p.d.f. f ( x ). We define its mean or expectation as E ( X ) ≡ Z ∞∞ xf ( x ) dx . We say that E ( X ) exists in this case iff Z ∞∞  x  f ( x ) dx < ∞ . Note that if there are numbers a , b such that∞ < a < b < ∞ and Pr ( a ≤ X ≤ b ) = 1, then E ( X ) exists and a ≤ E ( X ) ≤ b . 5 / 41 Example (The Uniform ( a , b ) Distribution) Let X ∼ U ( a , b ). Then f ( x ) = 1 b a , and hence E ( X ) = Z b a x 1 b a dx = 1 b a Z b a xdx = 1 b a x 2 2 b x = a = b 2 a 2 2( b a ) = b + a 2 . 6 / 41 The Expectation of a function of a r.v. If X is a r.v. and Y = r ( X ), then Y is also a r.v. and we can calculate E ( Y ). It is not necessary to determine the probability distribution of Y to compute it. Continuous Case X is a continuous X r.v., and we say E ( Y ) exists iff R ∞∞  r ( x )  f ( x ) dx < ∞ , and then we calculate E ( Y ) = Z ∞∞ r ( x ) f ( x ) dx . Discrete Case Replace the integrals by sums, i.e., E ( Y ) = X x r ( x ) f ( x ) . 7 / 41 Example (Expected Project Delay Penalty) Suppose X is the random variable representing the duration in months of a project. Suppose X ∼ U (5 , 10). Suppose the company in charge of the project agreed to finish the project in 7 months, and a penalty of a 2 million dollars will be paid by this company for each additional month it takes to finish. What is the expected penalty? We can define the penalty as a new random variable, which is a function of X . Specifically, we define Y = ( 2( Y 7) if Y ≥ 7, otherwise....
View
Full
Document
This note was uploaded on 01/02/2010 for the course APSC 116 taught by Professor Tommazzuchi during the Fall '08 term at GWU.
 Fall '08
 TomMazzuchi

Click to edit the document details