Ch5 Transformation of Random Variables

# Ch5 Transformation of Random Variables - Stat1801...

This preview shows pages 1–5. Sign up to view the full content.

Stat1801 Probability and Statistics: Foundations of Actuarial Science Fall 2010-2011 Chapter V Transformation of Random Variables § 5.1 Distribution of Functions of A Single Random Variable In general, functions of random variables are random variables. Example 5.1 1. If , then ( 1 , 0 ~ N Z ) ( ) 2 1 , 2 1 ~ 2 Γ = Z Y , ( ) 2 , ~ σμ μσ N Z X + = . 2. If () λ α , ~ Γ X , then Γ = a aX Y , ~ . 3. If , then ( 1 , 0 ~ U X ) Exp X Y ~ ln 1 = . Theorem Let X be a continuous random variable on distributed on a space S with pdf ( ) x f X . Let where g is a function such that exists. Then the pdf of Y can be ( X g Y = ) 1 g obtained by () () 1 1 y g dy d y g f y f X Y = , ( ) S g y . Proof Let be the distribution function of X . The distribution function of Y is x F X () ( ) ( ) ( ) y X g P y Y P y F Y = = . If g is a strictly increasing function, then () ( ) ( ) ( ) ( ) y g F y g X P y F X Y 1 1 = = ( ) y g F dy d y f X Y 1 = y g dy d y g f X 1 1 = . p. 91

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Stat1801 Probability and Statistics: Foundations of Actuarial Science Fall 2010-2011 On the other hand, if g is a strictly decreasing function, then () ( ) ( ) ( ) ( ) y g F y g X P y F X Y 1 1 1 = = () y g F dy d y f X Y 1 = y g dy d y g f X 1 1 = . Since y g dy d 1 is positive when g is increasing, and is negative when g is decreasing, the pdf of Y can be expressed as () () 1 1 y g dy d y g f y f X Y = , ( ) S g y . Example 5.2 Let and ( 1 , 0 ~ U X ) X X g Y log 1 λ = = . 1 = x f X , 1 0 < < x 1 , 0 = S , = , 0 S g Since y e x x y x g = = log 1 , the inverse function exists. The pdf of Y is y e y g = 1 y y X Y e dy d e f y f = y y e e = × = 1 , . 0 > y Therefore Exp X Y ~ log 1 = . p. 92
Stat1801 Probability and Statistics: Foundations of Actuarial Science Fall 2010-2011 Example 5.3 Let and . ( 1 , 0 ~ N X ) () X Y Φ = = , S , () ( ) 1 , 0 = Φ S Since Φ is a one-one function, 1 Φ exists. The pdf of Y is given by () () 1 1 y dy d y f y f X Y Φ Φ = y y 1 1 ' 1 Φ Φ Φ = φ , 1 = ( ) 1 , 0 y (as x x = Φ ') Hence . 1 , 0 ~ U Y Example 5.4 ( Log-normal Distribution ) Let ( ) 2 , ~ σ μ N X and . X e X g Y = = = 2 2 2 2 exp 2 1 πσ x x f X , < < x = , S , = , 0 S g Since , the inverse x y y e y x g x log = = = ( ) y y g log 1 = exists. The pdf of Y is given by log log y dy d y f y f X Y = ( ) = 2 2 2 2 log exp 2 1 y y , . 0 > y It is called the log-normal distribution which is commonly used to model multiplicative product of random quantities such as return rate on stock investment. p. 93

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Stat1801 Probability and Statistics: Foundations of Actuarial Science Fall 2010-2011 Example 5.5 Let and . ( 1 , 0 ~ N X ) 2 X Y = () () 2 2 2 1 x X e x x f = = π φ , < < x () = , S , () ( ) = , 0 S g Since y x y x y x g ± = = = 2 does not have a unique solution, does not exists. We cannot use the formula. y g 1 We may start from the distribution function. Consider () ( ) ( ) y X P y Y P y F Y = = 2 ( ) y X y P = ( ) 1 2 Φ = y ' 2 ' y dy d y y F y f Y Y Φ = = 2 1 2 1 exp 2 1 2 2 y y = 2 2 1 2 1 2 1 y e y = 2 1 2 1 2 1 2 1 2 1 y e y Γ = , . 0 > y Hence 2 1 , 2 1 ~ 2 Γ = X Y .
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 02/01/2012 for the course STAT 1801 taught by Professor Mrchung during the Fall '10 term at HKU.

### Page1 / 16

Ch5 Transformation of Random Variables - Stat1801...

This preview shows document pages 1 - 5. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online